Statistics
| Branch: | Tag: | Revision:

root / NEWS @ 1a732a74

History | View | Annotate | Download (101.5 kB)

1
News
2
====
3

    
4

    
5
Version 2.8.0 beta1
6
-------------------
7

    
8
*(unreleased)*
9

    
10
- The minimum Python version needed to run Ganeti is now 2.6.
11
- The :doc:`Remote API <rapi>` daemon now supports a command line flag
12
  to always require authentication, ``--require-authentication``. It can
13
  be specified in ``$sysconfdir/default/ganeti``.
14
- A new cluster attribute 'enabled_disk_templates' is introduced. It will
15
  be used to manage the disk templates to be used by instances in the cluster.
16
  Initially, it will be set to a list that includes plain, drbd, if they were
17
  enabled by specifying a volume group name, and file and sharedfile, if those
18
  were enabled at configure time. Additionally, it will include all disk
19
  templates that are currently used by instances. The order of disk templates
20
  will be based on Ganeti's history of supporting them. In the future, the
21
  first entry of the list will be used as a default disk template on instance
22
  creation.
23
- ``cfgupgrade`` now supports a ``--downgrade`` option to bring the
24
  configuration back to the previous stable version.
25

    
26

    
27
Version 2.7.0 rc1
28
-----------------
29

    
30
*(unreleased)*
31

    
32
Incompatible/important changes
33
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
34

    
35
- Instance policies for disk size were documented to be on a per-disk
36
  basis, but hail applied them to the sum of all disks. This has been
37
  fixed.
38
- ``hbal`` will now exit with status 0 if, during job execution over
39
  LUXI, early exit has been requested and all jobs are successful;
40
  before, exit status 1 was used, which cannot be differentiated from
41
  "job error" case
42
- Compatibility with newer versions of rbd has been fixed
43
- ``gnt-instance batch-create`` has been changed to use the bulk create
44
  opcode from Ganeti. This lead to incompatible changes in the format of
45
  the JSON file. It's now not a custom dict anymore but a dict
46
  compatible with the ``OpInstanceCreate`` opcode.
47
- Parent directories for file storage need to be listed in
48
  ``$sysconfdir/ganeti/file-storage-paths`` now. ``cfgupgrade`` will
49
  write the file automatically based on old configuration values, but it
50
  can not distribute it across all nodes and the file contents should be
51
  verified. Use ``gnt-cluster copyfile
52
  $sysconfdir/ganeti/file-storage-paths`` once the cluster has been
53
  upgraded. The reason for requiring this list of paths now is that
54
  before it would have been possible to inject new paths via RPC,
55
  allowing files to be created in arbitrary locations. The RPC protocol
56
  is protected using SSL/X.509 certificates, but as a design principle
57
  Ganeti does not permit arbitrary paths to be passed.
58
- The parsing of the variants file for OSes (see
59
  :manpage:`ganeti-os-interface(7)`) has been slightly changed: now empty
60
  lines and comment lines (starting with ``#``) are ignored for better
61
  readability.
62
- The ``setup-ssh`` tool added in Ganeti 2.2 has been replaced and is no
63
  longer available. ``gnt-node add`` now invokes a new tool on the
64
  destination node, named ``prepare-node-join``, to configure the SSH
65
  daemon. Paramiko is no longer necessary to configure nodes' SSH
66
  daemons via ``gnt-node add``.
67
- Draining (``gnt-cluster queue drain``) and un-draining the job queue
68
  (``gnt-cluster queue undrain``) now affects all nodes in a cluster and
69
  the flag is not reset after a master failover.
70
- Python 2.4 has *not* been tested with this release. Using 2.6 or above
71
  is recommended. 2.6 will be mandatory from the 2.8 series.
72

    
73

    
74
New features
75
~~~~~~~~~~~~
76

    
77
- New network management functionality to support automatic allocation
78
  of IP addresses and managing of network parameters. See
79
  :manpage:`gnt-network(8)` for more details.
80
- New external storage backend, to allow managing arbitrary storage
81
  systems external to the cluster. See
82
  :manpage:`ganeti-extstorage-interface(7)`.
83
- New ``exclusive-storage`` node parameter added, restricted to
84
  nodegroup level. When it's set to true, physical disks are assigned in
85
  an exclusive fashion to instances, as documented in :doc:`Partitioned
86
  Ganeti <design-partitioned>`.  Currently, only instances using the
87
  ``plain`` disk template are supported.
88
- The KVM hypervisor has been updated with many new hypervisor
89
  parameters, including a generic one for passing arbitrary command line
90
  values. See a complete list in :manpage:`gnt-instance(8)`. It is now
91
  compatible up to qemu 1.4.
92
- A new tool, called ``mon-collector``, is the stand-alone executor of
93
  the data collectors for a monitoring system. As of this version, it
94
  just includes the DRBD data collector, that can be executed by calling
95
  ``mon-collector`` using the ``drbd`` parameter. See
96
  :manpage:`mon-collector(7)`.
97
- A new user option, :pyeval:`rapi.RAPI_ACCESS_READ`, has been added
98
  for RAPI users. It allows granting permissions to query for
99
  information to a specific user without giving
100
  :pyeval:`rapi.RAPI_ACCESS_WRITE` permissions.
101
- A new tool named ``node-cleanup`` has been added. It cleans remains of
102
  a cluster from a machine by stopping all daemons, removing
103
  certificates and ssconf files. Unless the ``--no-backup`` option is
104
  given, copies of the certificates are made.
105
- Instance creations now support the use of opportunistic locking,
106
  potentially speeding up the (parallel) creation of multiple instances.
107
  This feature is currently only available via the :doc:`RAPI
108
  <rapi>` interface and when an instance allocator is used. If the
109
  ``opportunistic_locking`` parameter is set the opcode will try to
110
  acquire as many locks as possible, but will not wait for any locks
111
  held by other opcodes. If not enough resources can be found to
112
  allocate the instance, the temporary error code
113
  :pyeval:`errors.ECODE_TEMP_NORES` is returned. The operation can be
114
  retried thereafter, with or without opportunistic locking.
115
- New experimental linux-ha resource scripts.
116
- Restricted-commands support: ganeti can now be asked (via command line
117
  or rapi) to perform commands on a node. These are passed via ganeti
118
  RPC rather than ssh. This functionality is restricted to commands
119
  specified on the ``$sysconfdir/ganeti/restricted-commands`` for security
120
  reasons. The file is not copied automatically.
121

    
122

    
123
Misc changes
124
~~~~~~~~~~~~
125

    
126
- Diskless instances are now externally mirrored (Issue 237). This for
127
  now has only been tested in conjunction with explicit target nodes for
128
  migration/failover.
129
- Queries not needing locks or RPC access to the node can now be
130
  performed by the confd daemon, making them independent from jobs, and
131
  thus faster to execute. This is selectable at configure time.
132
- The functionality for allocating multiple instances at once has been
133
  overhauled and is now also available through :doc:`RAPI <rapi>`.
134

    
135

    
136
Since beta3:
137

    
138
- Fix kvm compatibility with qemu 1.4 (Issue 389)
139
- Documentation updates (admin guide, upgrade notes, install
140
  instructions) (Issue 372)
141
- Fix gnt-group list nodes and instances count (Issue 436)
142
- Fix compilation without non-mandatory libraries (Issue 441)
143
- Fix xen-hvm hypervisor forcing nics to type 'ioemu' (Issue 247)
144
- Make confd logging more verbose at INFO level (Issue 435)
145
- Improve "networks" documentation in :manpage:`gnt-instance(8)`
146
- Fix failure path for instance storage type conversion (Issue 229)
147
- Update htools text backend documentation
148
- Improve the renew-crypto section of :manpage:`gnt-cluster(8)`
149
- Disable inter-cluster instance move for file-based instances, because
150
  it is dependant on instance export, which is not supported for
151
  file-based instances. (Issue 414)
152

    
153

    
154
Version 2.7.0 beta3
155
-------------------
156

    
157
*(Released Mon, 22 Apr 2013)*
158

    
159
This was the third beta release of the 2.7 series. Since beta2:
160

    
161
- Fix hail to verify disk instance policies on a per-disk basis (Issue 418).
162
- Fix data loss on wrong usage of ``gnt-instance move``
163
- Properly export errors in confd-based job queries
164
- Add ``users-setup`` tool
165
- Fix iallocator protocol to report 0 as a disk size for diskless
166
  instances. This avoids hail breaking when a diskless instance is
167
  present.
168
- Fix job queue directory permission problem that made confd job queries
169
  fail. This requires running an ``ensure-dirs --full-run`` on upgrade
170
  for access to archived jobs (Issue 406).
171
- Limit the sizes of networks supported by ``gnt-network`` to something
172
  between a ``/16`` and a ``/30`` to prevent memory bloat and crashes.
173
- Fix bugs in instance disk template conversion
174
- Fix GHC 7 compatibility
175
- Fix ``burnin`` install path (Issue 426).
176
- Allow very small disk grows (Issue 347).
177
- Fix a ``ganeti-noded`` memory bloat introduced in 2.5, by making sure
178
  that noded doesn't import masterd code (Issue 419).
179
- Make sure the default metavg at cluster init is the same as the vg, if
180
  unspecified (Issue 358).
181
- Fix cleanup of partially created disks (part of Issue 416)
182

    
183

    
184
Version 2.7.0 beta2
185
-------------------
186

    
187
*(Released Tue, 2 Apr 2013)*
188

    
189
This was the second beta release of the 2.7 series. Since beta1:
190

    
191
- Networks no longer have a "type" slot, since this information was
192
  unused in Ganeti: instead of it tags should be used.
193
- The rapi client now has a ``target_node`` option to MigrateInstance.
194
- Fix early exit return code for hbal (Issue 386).
195
- Fix ``gnt-instance migrate/failover -n`` (Issue 396).
196
- Fix ``rbd showmapped`` output parsing (Issue 312).
197
- Networks are now referenced indexed by UUID, rather than name. This
198
  will require running cfgupgrade, from 2.7.0beta1, if networks are in
199
  use.
200
- The OS environment now includes network information.
201
- Deleting of a network is now disallowed if any instance nic is using
202
  it, to prevent dangling references.
203
- External storage is now documented in man pages.
204
- The exclusive_storage flag can now only be set at nodegroup level.
205
- Hbal can now submit an explicit priority with its jobs.
206
- Many network related locking fixes.
207
- Bump up the required pylint version to 0.25.1.
208
- Fix the ``no_remember`` option in RAPI client.
209
- Many ipolicy related tests, qa, and fixes.
210
- Many documentation improvements and fixes.
211
- Fix building with ``--disable-file-storage``.
212
- Fix ``-q`` option in htools, which was broken if passed more than
213
  once.
214
- Some haskell/python interaction improvements and fixes.
215
- Fix iallocator in case of missing LVM storage.
216
- Fix confd config load in case of ``--no-lvm-storage``.
217
- The confd/query functionality is now mentioned in the security
218
  documentation.
219

    
220

    
221
Version 2.7.0 beta1
222
-------------------
223

    
224
*(Released Wed, 6 Feb 2013)*
225

    
226
This was the first beta release of the 2.7 series. All important changes
227
are listed in the latest 2.7 entry.
228

    
229

    
230
Version 2.6.2
231
-------------
232

    
233
*(Released Fri, 21 Dec 2012)*
234

    
235
Important behaviour change: hbal won't rebalance anymore instances which
236
have the ``auto_balance`` attribute set to false. This was the intention
237
all along, but until now it only skipped those from the N+1 memory
238
reservation (DRBD-specific).
239

    
240
A significant number of bug fixes in this release:
241

    
242
- Fixed disk adoption interaction with ipolicy checks.
243
- Fixed networking issues when instances are started, stopped or
244
  migrated, by forcing the tap device's MAC prefix to "fe" (issue 217).
245
- Fixed the warning in cluster verify for shared storage instances not
246
  being redundant.
247
- Fixed removal of storage directory on shared file storage (issue 262).
248
- Fixed validation of LVM volume group name in OpClusterSetParams
249
  (``gnt-cluster modify``) (issue 285).
250
- Fixed runtime memory increases (``gnt-instance modify -m``).
251
- Fixed live migration under Xen's ``xl`` mode.
252
- Fixed ``gnt-instance console`` with ``xl``.
253
- Fixed building with newer Haskell compiler/libraries.
254
- Fixed PID file writing in Haskell daemons (confd); this prevents
255
  restart issues if confd was launched manually (outside of
256
  ``daemon-util``) while another copy of it was running
257
- Fixed a type error when doing live migrations with KVM (issue 297) and
258
  the error messages for failing migrations have been improved.
259
- Fixed opcode validation for the out-of-band commands (``gnt-node
260
  power``).
261
- Fixed a type error when unsetting OS hypervisor parameters (issue
262
  311); now it's possible to unset all OS-specific hypervisor
263
  parameters.
264
- Fixed the ``dry-run`` mode for many operations: verification of
265
  results was over-zealous but didn't take into account the ``dry-run``
266
  operation, resulting in "wrong" failures.
267
- Fixed bash completion in ``gnt-job list`` when the job queue has
268
  hundreds of entries; especially with older ``bash`` versions, this
269
  results in significant CPU usage.
270

    
271
And lastly, a few other improvements have been made:
272

    
273
- Added option to force master-failover without voting (issue 282).
274
- Clarified error message on lock conflict (issue 287).
275
- Logging of newly submitted jobs has been improved (issue 290).
276
- Hostname checks have been made uniform between instance rename and
277
  create (issue 291).
278
- The ``--submit`` option is now supported by ``gnt-debug delay``.
279
- Shutting down the master daemon by sending SIGTERM now stops it from
280
  processing jobs waiting for locks; instead, those jobs will be started
281
  once again after the master daemon is started the next time (issue
282
  296).
283
- Support for Xen's ``xl`` program has been improved (besides the fixes
284
  above).
285
- Reduced logging noise in the Haskell confd daemon (only show one log
286
  entry for each config reload, instead of two).
287
- Several man page updates and typo fixes.
288

    
289

    
290
Version 2.6.1
291
-------------
292

    
293
*(Released Fri, 12 Oct 2012)*
294

    
295
A small bugfix release. Among the bugs fixed:
296

    
297
- Fixed double use of ``PRIORITY_OPT`` in ``gnt-node migrate``, that
298
  made the command unusable.
299
- Commands that issue many jobs don't fail anymore just because some jobs
300
  take so long that other jobs are archived.
301
- Failures during ``gnt-instance reinstall`` are reflected by the exit
302
  status.
303
- Issue 190 fixed. Check for DRBD in cluster verify is enabled only when
304
  DRBD is enabled.
305
- When ``always_failover`` is set, ``--allow-failover`` is not required
306
  in migrate commands anymore.
307
- ``bash_completion`` works even if extglob is disabled.
308
- Fixed bug with locks that made failover for RDB-based instances fail.
309
- Fixed bug in non-mirrored instance allocation that made Ganeti choose
310
  a random node instead of one based on the allocator metric.
311
- Support for newer versions of pylint and pep8.
312
- Hail doesn't fail anymore when trying to add an instance of type
313
  ``file``, ``sharedfile`` or ``rbd``.
314
- Added new Makefile target to rebuild the whole distribution, so that
315
  all files are included.
316

    
317

    
318
Version 2.6.0
319
-------------
320

    
321
*(Released Fri, 27 Jul 2012)*
322

    
323

    
324
.. attention:: The ``LUXI`` protocol has been made more consistent
325
   regarding its handling of command arguments. This, however, leads to
326
   incompatibility issues with previous versions. Please ensure that you
327
   restart Ganeti daemons soon after the upgrade, otherwise most
328
   ``LUXI`` calls (job submission, setting/resetting the drain flag,
329
   pausing/resuming the watcher, cancelling and archiving jobs, querying
330
   the cluster configuration) will fail.
331

    
332

    
333
New features
334
~~~~~~~~~~~~
335

    
336
Instance run status
337
+++++++++++++++++++
338

    
339
The current ``admin_up`` field, which used to denote whether an instance
340
should be running or not, has been removed. Instead, ``admin_state`` is
341
introduced, with 3 possible values -- ``up``, ``down`` and ``offline``.
342

    
343
The rational behind this is that an instance being “down” can have
344
different meanings:
345

    
346
- it could be down during a reboot
347
- it could be temporarily be down for a reinstall
348
- or it could be down because it is deprecated and kept just for its
349
  disk
350

    
351
The previous Boolean state was making it difficult to do capacity
352
calculations: should Ganeti reserve memory for a down instance? Now, the
353
tri-state field makes it clear:
354

    
355
- in ``up`` and ``down`` state, all resources are reserved for the
356
  instance, and it can be at any time brought up if it is down
357
- in ``offline`` state, only disk space is reserved for it, but not
358
  memory or CPUs
359

    
360
The field can have an extra use: since the transition between ``up`` and
361
``down`` and vice-versus is done via ``gnt-instance start/stop``, but
362
transition between ``offline`` and ``down`` is done via ``gnt-instance
363
modify``, it is possible to given different rights to users. For
364
example, owners of an instance could be allowed to start/stop it, but
365
not transition it out of the offline state.
366

    
367
Instance policies and specs
368
+++++++++++++++++++++++++++
369

    
370
In previous Ganeti versions, an instance creation request was not
371
limited on the minimum size and on the maximum size just by the cluster
372
resources. As such, any policy could be implemented only in third-party
373
clients (RAPI clients, or shell wrappers over ``gnt-*``
374
tools). Furthermore, calculating cluster capacity via ``hspace`` again
375
required external input with regards to instance sizes.
376

    
377
In order to improve these workflows and to allow for example better
378
per-node group differentiation, we introduced instance specs, which
379
allow declaring:
380

    
381
- minimum instance disk size, disk count, memory size, cpu count
382
- maximum values for the above metrics
383
- and “standard” values (used in ``hspace`` to calculate the standard
384
  sized instances)
385

    
386
The minimum/maximum values can be also customised at node-group level,
387
for example allowing more powerful hardware to support bigger instance
388
memory sizes.
389

    
390
Beside the instance specs, there are a few other settings belonging to
391
the instance policy framework. It is possible now to customise, per
392
cluster and node-group:
393

    
394
- the list of allowed disk templates
395
- the maximum ratio of VCPUs per PCPUs (to control CPU oversubscription)
396
- the maximum ratio of instance to spindles (see below for more
397
  information) for local storage
398

    
399
All these together should allow all tools that talk to Ganeti to know
400
what are the ranges of allowed values for instances and the
401
over-subscription that is allowed.
402

    
403
For the VCPU/PCPU ratio, we already have the VCPU configuration from the
404
instance configuration, and the physical CPU configuration from the
405
node. For the spindle ratios however, we didn't track before these
406
values, so new parameters have been added:
407

    
408
- a new node parameter ``spindle_count``, defaults to 1, customisable at
409
  node group or node level
410
- at new backend parameter (for instances), ``spindle_use`` defaults to 1
411

    
412
Note that spindles in this context doesn't need to mean actual
413
mechanical hard-drives; it's just a relative number for both the node
414
I/O capacity and instance I/O consumption.
415

    
416
Instance migration behaviour
417
++++++++++++++++++++++++++++
418

    
419
While live-migration is in general desirable over failover, it is
420
possible that for some workloads it is actually worse, due to the
421
variable time of the “suspend” phase during live migration.
422

    
423
To allow the tools to work consistently over such instances (without
424
having to hard-code instance names), a new backend parameter
425
``always_failover`` has been added to control the migration/failover
426
behaviour. When set to True, all migration requests for an instance will
427
instead fall-back to failover.
428

    
429
Instance memory ballooning
430
++++++++++++++++++++++++++
431

    
432
Initial support for memory ballooning has been added. The memory for an
433
instance is no longer fixed (backend parameter ``memory``), but instead
434
can vary between minimum and maximum values (backend parameters
435
``minmem`` and ``maxmem``). Currently we only change an instance's
436
memory when:
437

    
438
- live migrating or failing over and instance and the target node
439
  doesn't have enough memory
440
- user requests changing the memory via ``gnt-instance modify
441
  --runtime-memory``
442

    
443
Instance CPU pinning
444
++++++++++++++++++++
445

    
446
In order to control the use of specific CPUs by instance, support for
447
controlling CPU pinning has been added for the Xen, HVM and LXC
448
hypervisors. This is controlled by a new hypervisor parameter
449
``cpu_mask``; details about possible values for this are in the
450
:manpage:`gnt-instance(8)`. Note that use of the most specific (precise
451
VCPU-to-CPU mapping) form will work well only when all nodes in your
452
cluster have the same amount of CPUs.
453

    
454
Disk parameters
455
+++++++++++++++
456

    
457
Another area in which Ganeti was not customisable were the parameters
458
used for storage configuration, e.g. how many stripes to use for LVM,
459
DRBD resync configuration, etc.
460

    
461
To improve this area, we've added disks parameters, which are
462
customisable at cluster and node group level, and which allow to
463
specify various parameters for disks (DRBD has the most parameters
464
currently), for example:
465

    
466
- DRBD resync algorithm and parameters (e.g. speed)
467
- the default VG for meta-data volumes for DRBD
468
- number of stripes for LVM (plain disk template)
469
- the RBD pool
470

    
471
These parameters can be modified via ``gnt-cluster modify -D …`` and
472
``gnt-group modify -D …``, and are used at either instance creation (in
473
case of LVM stripes, for example) or at disk “activation” time
474
(e.g. resync speed).
475

    
476
Rados block device support
477
++++++++++++++++++++++++++
478

    
479
A Rados (http://ceph.com/wiki/Rbd) storage backend has been added,
480
denoted by the ``rbd`` disk template type. This is considered
481
experimental, feedback is welcome. For details on configuring it, see
482
the :doc:`install` document and the :manpage:`gnt-cluster(8)` man page.
483

    
484
Master IP setup
485
+++++++++++++++
486

    
487
The existing master IP functionality works well only in simple setups (a
488
single network shared by all nodes); however, if nodes belong to
489
different networks, then the ``/32`` setup and lack of routing
490
information is not enough.
491

    
492
To allow the master IP to function well in more complex cases, the
493
system was reworked as follows:
494

    
495
- a master IP netmask setting has been added
496
- the master IP activation/turn-down code was moved from the node daemon
497
  to a separate script
498
- whether to run the Ganeti-supplied master IP script or a user-supplied
499
  on is a ``gnt-cluster init`` setting
500

    
501
Details about the location of the standard and custom setup scripts are
502
in the man page :manpage:`gnt-cluster(8)`; for information about the
503
setup script protocol, look at the Ganeti-supplied script.
504

    
505
SPICE support
506
+++++++++++++
507

    
508
The `SPICE <http://www.linux-kvm.org/page/SPICE>`_ support has been
509
improved.
510

    
511
It is now possible to use TLS-protected connections, and when renewing
512
or changing the cluster certificates (via ``gnt-cluster renew-crypto``,
513
it is now possible to specify spice or spice CA certificates. Also, it
514
is possible to configure a password for SPICE sessions via the
515
hypervisor parameter ``spice_password_file``.
516

    
517
There are also new parameters to control the compression and streaming
518
options (e.g. ``spice_image_compression``, ``spice_streaming_video``,
519
etc.). For details, see the man page :manpage:`gnt-instance(8)` and look
520
for the spice parameters.
521

    
522
Lastly, it is now possible to see the SPICE connection information via
523
``gnt-instance console``.
524

    
525
OVF converter
526
+++++++++++++
527

    
528
A new tool (``tools/ovfconverter``) has been added that supports
529
conversion between Ganeti and the `Open Virtualization Format
530
<http://en.wikipedia.org/wiki/Open_Virtualization_Format>`_ (both to and
531
from).
532

    
533
This relies on the ``qemu-img`` tool to convert the disk formats, so the
534
actual compatibility with other virtualization solutions depends on it.
535

    
536
Confd daemon changes
537
++++++++++++++++++++
538

    
539
The configuration query daemon (``ganeti-confd``) is now optional, and
540
has been rewritten in Haskell; whether to use the daemon at all, use the
541
Python (default) or the Haskell version is selectable at configure time
542
via the ``--enable-confd`` parameter, which can take one of the
543
``haskell``, ``python`` or ``no`` values. If not used, disabling the
544
daemon will result in a smaller footprint; for larger systems, we
545
welcome feedback on the Haskell version which might become the default
546
in future versions.
547

    
548
If you want to use ``gnt-node list-drbd`` you need to have the Haskell
549
daemon running. The Python version doesn't implement the new call.
550

    
551

    
552
User interface changes
553
~~~~~~~~~~~~~~~~~~~~~~
554

    
555
We have replaced the ``--disks`` option of ``gnt-instance
556
replace-disks`` with a more flexible ``--disk`` option, which allows
557
adding and removing disks at arbitrary indices (Issue 188). Furthermore,
558
disk size and mode can be changed upon recreation (via ``gnt-instance
559
recreate-disks``, which accepts the same ``--disk`` option).
560

    
561
As many people are used to a ``show`` command, we have added that as an
562
alias to ``info`` on all ``gnt-*`` commands.
563

    
564
The ``gnt-instance grow-disk`` command has a new mode in which it can
565
accept the target size of the disk, instead of the delta; this can be
566
more safe since two runs in absolute mode will be idempotent, and
567
sometimes it's also easier to specify the desired size directly.
568

    
569
Also the handling of instances with regard to offline secondaries has
570
been improved. Instance operations should not fail because one of it's
571
secondary nodes is offline, even though it's safe to proceed.
572

    
573
A new command ``list-drbd`` has been added to the ``gnt-node`` script to
574
support debugging of DRBD issues on nodes. It provides a mapping of DRBD
575
minors to instance name.
576

    
577
API changes
578
~~~~~~~~~~~
579

    
580
RAPI coverage has improved, with (for example) new resources for
581
recreate-disks, node power-cycle, etc.
582

    
583
Compatibility
584
~~~~~~~~~~~~~
585

    
586
There is partial support for ``xl`` in the Xen hypervisor; feedback is
587
welcome.
588

    
589
Python 2.7 is better supported, and after Ganeti 2.6 we will investigate
590
whether to still support Python 2.4 or move to Python 2.6 as minimum
591
required version.
592

    
593
Support for Fedora has been slightly improved; the provided example
594
init.d script should work better on it and the INSTALL file should
595
document the needed dependencies.
596

    
597
Internal changes
598
~~~~~~~~~~~~~~~~
599

    
600
The deprecated ``QueryLocks`` LUXI request has been removed. Use
601
``Query(what=QR_LOCK, ...)`` instead.
602

    
603
The LUXI requests :pyeval:`luxi.REQ_QUERY_JOBS`,
604
:pyeval:`luxi.REQ_QUERY_INSTANCES`, :pyeval:`luxi.REQ_QUERY_NODES`,
605
:pyeval:`luxi.REQ_QUERY_GROUPS`, :pyeval:`luxi.REQ_QUERY_EXPORTS` and
606
:pyeval:`luxi.REQ_QUERY_TAGS` are deprecated and will be removed in a
607
future version. :pyeval:`luxi.REQ_QUERY` should be used instead.
608

    
609
RAPI client: ``CertificateError`` now derives from
610
``GanetiApiError``. This should make it more easy to handle Ganeti
611
errors.
612

    
613
Deprecation warnings due to PyCrypto/paramiko import in
614
``tools/setup-ssh`` have been silenced, as usually they are safe; please
615
make sure to run an up-to-date paramiko version, if you use this tool.
616

    
617
The QA scripts now depend on Python 2.5 or above (the main code base
618
still works with Python 2.4).
619

    
620
The configuration file (``config.data``) is now written without
621
indentation for performance reasons; if you want to edit it, it can be
622
re-formatted via ``tools/fmtjson``.
623

    
624
A number of bugs has been fixed in the cluster merge tool.
625

    
626
``x509`` certification verification (used in import-export) has been
627
changed to allow the same clock skew as permitted by the cluster
628
verification. This will remove some rare but hard to diagnose errors in
629
import-export.
630

    
631

    
632
Version 2.6.0 rc4
633
-----------------
634

    
635
*(Released Thu, 19 Jul 2012)*
636

    
637
Very few changes from rc4 to the final release, only bugfixes:
638

    
639
- integrated fixes from release 2.5.2 (fix general boot flag for KVM
640
  instance, fix CDROM booting for KVM instances)
641
- fixed node group modification of node parameters
642
- fixed issue in LUClusterVerifyGroup with multi-group clusters
643
- fixed generation of bash completion to ensure a stable ordering
644
- fixed a few typos
645

    
646

    
647
Version 2.6.0 rc3
648
-----------------
649

    
650
*(Released Fri, 13 Jul 2012)*
651

    
652
Third release candidate for 2.6. The following changes were done from
653
rc3 to rc4:
654

    
655
- Fixed ``UpgradeConfig`` w.r.t. to disk parameters on disk objects.
656
- Fixed an inconsistency in the LUXI protocol with the provided
657
  arguments (NOT backwards compatible)
658
- Fixed a bug with node groups ipolicy where ``min`` was greater than
659
  the cluster ``std`` value
660
- Implemented a new ``gnt-node list-drbd`` call to list DRBD minors for
661
  easier instance debugging on nodes (requires ``hconfd`` to work)
662

    
663

    
664
Version 2.6.0 rc2
665
-----------------
666

    
667
*(Released Tue, 03 Jul 2012)*
668

    
669
Second release candidate for 2.6. The following changes were done from
670
rc2 to rc3:
671

    
672
- Fixed ``gnt-cluster verify`` regarding ``master-ip-script`` on non
673
  master candidates
674
- Fixed a RAPI regression on missing beparams/memory
675
- Fixed redistribution of files on offline nodes
676
- Added possibility to run activate-disks even though secondaries are
677
  offline. With this change it relaxes also the strictness on some other
678
  commands which use activate disks internally:
679
  * ``gnt-instance start|reboot|rename|backup|export``
680
- Made it possible to remove safely an instance if its secondaries are
681
  offline
682
- Made it possible to reinstall even though secondaries are offline
683

    
684

    
685
Version 2.6.0 rc1
686
-----------------
687

    
688
*(Released Mon, 25 Jun 2012)*
689

    
690
First release candidate for 2.6. The following changes were done from
691
rc1 to rc2:
692

    
693
- Fixed bugs with disk parameters and ``rbd`` templates as well as
694
  ``instance_os_add``
695
- Made ``gnt-instance modify`` more consistent regarding new NIC/Disk
696
  behaviour. It supports now the modify operation
697
- ``hcheck`` implemented to analyze cluster health and possibility of
698
  improving health by rebalance
699
- ``hbal`` has been improved in dealing with split instances
700

    
701

    
702
Version 2.6.0 beta2
703
-------------------
704

    
705
*(Released Mon, 11 Jun 2012)*
706

    
707
Second beta release of 2.6. The following changes were done from beta2
708
to rc1:
709

    
710
- Fixed ``daemon-util`` with non-root user models
711
- Fixed creation of plain instances with ``--no-wait-for-sync``
712
- Fix wrong iv_names when running ``cfgupgrade``
713
- Export more information in RAPI group queries
714
- Fixed bug when changing instance network interfaces
715
- Extended burnin to do NIC changes
716
- query: Added ``<``, ``>``, ``<=``, ``>=`` comparison operators
717
- Changed default for DRBD barriers
718
- Fixed DRBD error reporting for syncer rate
719
- Verify the options on disk parameters
720

    
721
And of course various fixes to documentation and improved unittests and
722
QA.
723

    
724

    
725
Version 2.6.0 beta1
726
-------------------
727

    
728
*(Released Wed, 23 May 2012)*
729

    
730
First beta release of 2.6. The following changes were done from beta1 to
731
beta2:
732

    
733
- integrated patch for distributions without ``start-stop-daemon``
734
- adapted example init.d script to work on Fedora
735
- fixed log handling in Haskell daemons
736
- adapted checks in the watcher for pycurl linked against libnss
737
- add partial support for ``xl`` instead of ``xm`` for Xen
738
- fixed a type issue in cluster verification
739
- fixed ssconf handling in the Haskell code (was breaking confd in IPv6
740
  clusters)
741

    
742
Plus integrated fixes from the 2.5 branch:
743

    
744
- fixed ``kvm-ifup`` to use ``/bin/bash``
745
- fixed parallel build failures
746
- KVM live migration when using a custom keymap
747

    
748

    
749
Version 2.5.2
750
-------------
751

    
752
*(Released Tue, 24 Jul 2012)*
753

    
754
A small bugfix release, with no new features:
755

    
756
- fixed bash-isms in kvm-ifup, for compatibility with systems which use a
757
  different default shell (e.g. Debian, Ubuntu)
758
- fixed KVM startup and live migration with a custom keymap (fixes Issue
759
  243 and Debian bug #650664)
760
- fixed compatibility with KVM versions that don't support multiple boot
761
  devices (fixes Issue 230 and Debian bug #624256)
762

    
763
Additionally, a few fixes were done to the build system (fixed parallel
764
build failures) and to the unittests (fixed race condition in test for
765
FileID functions, and the default enable/disable mode for QA test is now
766
customisable).
767

    
768

    
769
Version 2.5.1
770
-------------
771

    
772
*(Released Fri, 11 May 2012)*
773

    
774
A small bugfix release.
775

    
776
The main issues solved are on the topic of compatibility with newer LVM
777
releases:
778

    
779
- fixed parsing of ``lv_attr`` field
780
- adapted to new ``vgreduce --removemissing`` behaviour where sometimes
781
  the ``--force`` flag is needed
782

    
783
Also on the topic of compatibility, ``tools/lvmstrap`` has been changed
784
to accept kernel 3.x too (was hardcoded to 2.6.*).
785

    
786
A regression present in 2.5.0 that broke handling (in the gnt-* scripts)
787
of hook results and that also made display of other errors suboptimal
788
was fixed; the code behaves now like 2.4 and earlier.
789

    
790
Another change in 2.5, the cleanup of the OS scripts environment, is too
791
aggressive: it removed even the ``PATH`` variable, which requires the OS
792
scripts to *always* need to export it. Since this is a bit too strict,
793
we now export a minimal PATH, the same that we export for hooks.
794

    
795
The fix for issue 201 (Preserve bridge MTU in KVM ifup script) was
796
integrated into this release.
797

    
798
Finally, a few other miscellaneous changes were done (no new features,
799
just small improvements):
800

    
801
- Fix ``gnt-group --help`` display
802
- Fix hardcoded Xen kernel path
803
- Fix grow-disk handling of invalid units
804
- Update synopsis for ``gnt-cluster repair-disk-sizes``
805
- Accept both PUT and POST in noded (makes future upgrade to 2.6 easier)
806

    
807

    
808
Version 2.5.0
809
-------------
810

    
811
*(Released Thu, 12 Apr 2012)*
812

    
813
Incompatible/important changes and bugfixes
814
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
815

    
816
- The default of the ``/2/instances/[instance_name]/rename`` RAPI
817
  resource's ``ip_check`` parameter changed from ``True`` to ``False``
818
  to match the underlying LUXI interface.
819
- The ``/2/nodes/[node_name]/evacuate`` RAPI resource was changed to use
820
  body parameters, see :doc:`RAPI documentation <rapi>`. The server does
821
  not maintain backwards-compatibility as the underlying operation
822
  changed in an incompatible way. The RAPI client can talk to old
823
  servers, but it needs to be told so as the return value changed.
824
- When creating file-based instances via RAPI, the ``file_driver``
825
  parameter no longer defaults to ``loop`` and must be specified.
826
- The deprecated ``bridge`` NIC parameter is no longer supported. Use
827
  ``link`` instead.
828
- Support for the undocumented and deprecated RAPI instance creation
829
  request format version 0 has been dropped. Use version 1, supported
830
  since Ganeti 2.1.3 and :doc:`documented <rapi>`, instead.
831
- Pyparsing 1.4.6 or above is required, see :doc:`installation
832
  documentation <install>`.
833
- The "cluster-verify" hooks are now executed per group by the
834
  ``OP_CLUSTER_VERIFY_GROUP`` opcode. This maintains the same behavior
835
  if you just run ``gnt-cluster verify``, which generates one opcode per
836
  group.
837
- The environment as passed to the OS scripts is cleared, and thus no
838
  environment variables defined in the node daemon's environment will be
839
  inherited by the scripts.
840
- The :doc:`iallocator <iallocator>` mode ``multi-evacuate`` has been
841
  deprecated.
842
- :doc:`New iallocator modes <design-multi-reloc>` have been added to
843
  support operations involving multiple node groups.
844
- Offline nodes are ignored when failing over an instance.
845
- Support for KVM version 1.0, which changed the version reporting format
846
  from 3 to 2 digits.
847
- TCP/IP ports used by DRBD disks are returned to a pool upon instance
848
  removal.
849
- ``Makefile`` is now compatible with Automake 1.11.2
850
- Includes all bugfixes made in the 2.4 series
851

    
852
New features
853
~~~~~~~~~~~~
854

    
855
- The ganeti-htools project has been merged into the ganeti-core source
856
  tree and will be built as part of Ganeti (see :doc:`install-quick`).
857
- Implemented support for :doc:`shared storage <design-shared-storage>`.
858
- Add support for disks larger than 2 TB in ``lvmstrap`` by supporting
859
  GPT-style partition tables (requires `parted
860
  <http://www.gnu.org/s/parted/>`_).
861
- Added support for floppy drive and 2nd CD-ROM drive in KVM hypervisor.
862
- Allowed adding tags on instance creation.
863
- Export instance tags to hooks (``INSTANCE_TAGS``, see :doc:`hooks`)
864
- Allow instances to be started in a paused state, enabling the user to
865
  see the complete console output on boot using the console.
866
- Added new hypervisor flag to control default reboot behaviour
867
  (``reboot_behavior``).
868
- Added support for KVM keymaps (hypervisor parameter ``keymap``).
869
- Improved out-of-band management support:
870

    
871
  - Added ``gnt-node health`` command reporting the health status of
872
    nodes.
873
  - Added ``gnt-node power`` command to manage power status of nodes.
874
  - Added command for emergency power-off (EPO), ``gnt-cluster epo``.
875

    
876
- Instance migration can fall back to failover if instance is not
877
  running.
878
- Filters can be used when listing nodes, instances, groups and locks;
879
  see :manpage:`ganeti(7)` manpage.
880
- Added post-execution status as variables to :doc:`hooks <hooks>`
881
  environment.
882
- Instance tags are exported/imported together with the instance.
883
- When given an explicit job ID, ``gnt-job info`` will work for archived
884
  jobs.
885
- Jobs can define dependencies on other jobs (not yet supported via
886
  RAPI or command line, but used by internal commands and usable via
887
  LUXI).
888

    
889
  - Lock monitor (``gnt-debug locks``) shows jobs waiting for
890
    dependencies.
891

    
892
- Instance failover is now available as a RAPI resource
893
  (``/2/instances/[instance_name]/failover``).
894
- ``gnt-instance info`` defaults to static information if primary node
895
  is offline.
896
- Opcodes have a new ``comment`` attribute.
897
- Added basic SPICE support to KVM hypervisor.
898
- ``tools/ganeti-listrunner`` allows passing of arguments to executable.
899

    
900
Node group improvements
901
~~~~~~~~~~~~~~~~~~~~~~~
902

    
903
- ``gnt-cluster verify`` has been modified to check groups separately,
904
  thereby improving performance.
905
- Node group support has been added to ``gnt-cluster verify-disks``,
906
  which now operates per node group.
907
- Watcher has been changed to work better with node groups.
908

    
909
  - One process and state file per node group.
910
  - Slow watcher in one group doesn't block other group's watcher.
911

    
912
- Added new command, ``gnt-group evacuate``, to move all instances in a
913
  node group to other groups.
914
- Added ``gnt-instance change-group`` to move an instance to another
915
  node group.
916
- ``gnt-cluster command`` and ``gnt-cluster copyfile`` now support
917
  per-group operations.
918
- Node groups can be tagged.
919
- Some operations switch from an exclusive to a shared lock as soon as
920
  possible.
921
- Instance's primary and secondary nodes' groups are now available as
922
  query fields (``pnode.group``, ``pnode.group.uuid``, ``snodes.group``
923
  and ``snodes.group.uuid``).
924

    
925
Misc
926
~~~~
927

    
928
- Numerous updates to documentation and manpages.
929

    
930
  - :doc:`RAPI <rapi>` documentation now has detailed parameter
931
    descriptions.
932
  - Some opcode/job results are now also documented, see :doc:`RAPI
933
    <rapi>`.
934

    
935
- A lockset's internal lock is now also visible in lock monitor.
936
- Log messages from job queue workers now contain information about the
937
  opcode they're processing.
938
- ``gnt-instance console`` no longer requires the instance lock.
939
- A short delay when waiting for job changes reduces the number of LUXI
940
  requests significantly.
941
- DRBD metadata volumes are overwritten with zeros during disk creation.
942
- Out-of-band commands no longer acquire the cluster lock in exclusive
943
  mode.
944
- ``devel/upload`` now uses correct permissions for directories.
945

    
946

    
947
Version 2.5.0 rc6
948
-----------------
949

    
950
*(Released Fri, 23 Mar 2012)*
951

    
952
This was the sixth release candidate of the 2.5 series.
953

    
954

    
955
Version 2.5.0 rc5
956
-----------------
957

    
958
*(Released Mon, 9 Jan 2012)*
959

    
960
This was the fifth release candidate of the 2.5 series.
961

    
962

    
963
Version 2.5.0 rc4
964
-----------------
965

    
966
*(Released Thu, 27 Oct 2011)*
967

    
968
This was the fourth release candidate of the 2.5 series.
969

    
970

    
971
Version 2.5.0 rc3
972
-----------------
973

    
974
*(Released Wed, 26 Oct 2011)*
975

    
976
This was the third release candidate of the 2.5 series.
977

    
978

    
979
Version 2.5.0 rc2
980
-----------------
981

    
982
*(Released Tue, 18 Oct 2011)*
983

    
984
This was the second release candidate of the 2.5 series.
985

    
986

    
987
Version 2.5.0 rc1
988
-----------------
989

    
990
*(Released Tue, 4 Oct 2011)*
991

    
992
This was the first release candidate of the 2.5 series.
993

    
994

    
995
Version 2.5.0 beta3
996
-------------------
997

    
998
*(Released Wed, 31 Aug 2011)*
999

    
1000
This was the third beta release of the 2.5 series.
1001

    
1002

    
1003
Version 2.5.0 beta2
1004
-------------------
1005

    
1006
*(Released Mon, 22 Aug 2011)*
1007

    
1008
This was the second beta release of the 2.5 series.
1009

    
1010

    
1011
Version 2.5.0 beta1
1012
-------------------
1013

    
1014
*(Released Fri, 12 Aug 2011)*
1015

    
1016
This was the first beta release of the 2.5 series.
1017

    
1018

    
1019
Version 2.4.5
1020
-------------
1021

    
1022
*(Released Thu, 27 Oct 2011)*
1023

    
1024
- Fixed bug when parsing command line parameter values ending in
1025
  backslash
1026
- Fixed assertion error after unclean master shutdown
1027
- Disable HTTP client pool for RPC, significantly reducing memory usage
1028
  of master daemon
1029
- Fixed queue archive creation with wrong permissions
1030

    
1031

    
1032
Version 2.4.4
1033
-------------
1034

    
1035
*(Released Tue, 23 Aug 2011)*
1036

    
1037
Small bug-fixes:
1038

    
1039
- Fixed documentation for importing with ``--src-dir`` option
1040
- Fixed a bug in ``ensure-dirs`` with queue/archive permissions
1041
- Fixed a parsing issue with DRBD 8.3.11 in the Linux kernel
1042

    
1043

    
1044
Version 2.4.3
1045
-------------
1046

    
1047
*(Released Fri, 5 Aug 2011)*
1048

    
1049
Many bug-fixes and a few small features:
1050

    
1051
- Fixed argument order in ``ReserveLV`` and ``ReserveMAC`` which caused
1052
  issues when you tried to add an instance with two MAC addresses in one
1053
  request
1054
- KVM: fixed per-instance stored UID value
1055
- KVM: configure bridged NICs at migration start
1056
- KVM: Fix a bug where instance will not start with never KVM versions
1057
  (>= 0.14)
1058
- Added OS search path to ``gnt-cluster info``
1059
- Fixed an issue with ``file_storage_dir`` where you were forced to
1060
  provide an absolute path, but the documentation states it is a
1061
  relative path, the documentation was right
1062
- Added a new parameter to instance stop/start called ``--no-remember``
1063
  that will make the state change to not be remembered
1064
- Implemented ``no_remember`` at RAPI level
1065
- Improved the documentation
1066
- Node evacuation: don't call IAllocator if node is already empty
1067
- Fixed bug in DRBD8 replace disks on current nodes
1068
- Fixed bug in recreate-disks for DRBD instances
1069
- Moved assertion checking locks in ``gnt-instance replace-disks``
1070
  causing it to abort with not owning the right locks for some situation
1071
- Job queue: Fixed potential race condition when cancelling queued jobs
1072
- Fixed off-by-one bug in job serial generation
1073
- ``gnt-node volumes``: Fix instance names
1074
- Fixed aliases in bash completion
1075
- Fixed a bug in reopening log files after being sent a SIGHUP
1076
- Added a flag to burnin to allow specifying VCPU count
1077
- Bugfixes to non-root Ganeti configuration
1078

    
1079

    
1080
Version 2.4.2
1081
-------------
1082

    
1083
*(Released Thu, 12 May 2011)*
1084

    
1085
Many bug-fixes and a few new small features:
1086

    
1087
- Fixed a bug related to log opening failures
1088
- Fixed a bug in instance listing with orphan instances
1089
- Fixed a bug which prevented resetting the cluster-level node parameter
1090
  ``oob_program`` to the default
1091
- Many fixes related to the ``cluster-merge`` tool
1092
- Fixed a race condition in the lock monitor, which caused failures
1093
  during (at least) creation of many instances in parallel
1094
- Improved output for gnt-job info
1095
- Removed the quiet flag on some ssh calls which prevented debugging
1096
  failures
1097
- Improved the N+1 failure messages in cluster verify by actually
1098
  showing the memory values (needed and available)
1099
- Increased lock attempt timeouts so that when executing long operations
1100
  (e.g. DRBD replace-disks) other jobs do not enter 'blocking acquire'
1101
  too early and thus prevent the use of the 'fair' mechanism
1102
- Changed instance query data (``gnt-instance info``) to not acquire
1103
  locks unless needed, thus allowing its use on locked instance if only
1104
  static information is asked for
1105
- Improved behaviour with filesystems that do not support rename on an
1106
  opened file
1107
- Fixed the behaviour of ``prealloc_wipe_disks`` cluster parameter which
1108
  kept locks on all nodes during the wipe, which is unneeded
1109
- Fixed ``gnt-watcher`` handling of errors during hooks execution
1110
- Fixed bug in ``prealloc_wipe_disks`` with small disk sizes (less than
1111
  10GiB) which caused the wipe to fail right at the end in some cases
1112
- Fixed master IP activation when doing master failover with no-voting
1113
- Fixed bug in ``gnt-node add --readd`` which allowed the re-adding of
1114
  the master node itself
1115
- Fixed potential data-loss in under disk full conditions, where Ganeti
1116
  wouldn't check correctly the return code and would consider
1117
  partially-written files 'correct'
1118
- Fixed bug related to multiple VGs and DRBD disk replacing
1119
- Added new disk parameter ``metavg`` that allows placement of the meta
1120
  device for DRBD in a different volume group
1121
- Fixed error handling in the node daemon when the system libc doesn't
1122
  have major number 6 (i.e. if ``libc.so.6`` is not the actual libc)
1123
- Fixed lock release during replace-disks, which kept cluster-wide locks
1124
  when doing disk replaces with an iallocator script
1125
- Added check for missing bridges in cluster verify
1126
- Handle EPIPE errors while writing to the terminal better, so that
1127
  piping the output to e.g. ``less`` doesn't cause a backtrace
1128
- Fixed rare case where a ^C during Luxi calls could have been
1129
  interpreted as server errors, instead of simply terminating
1130
- Fixed a race condition in LUGroupAssignNodes (``gnt-group
1131
  assign-nodes``)
1132
- Added a few more parameters to the KVM hypervisor, allowing a second
1133
  CDROM, custom disk type for CDROMs and a floppy image
1134
- Removed redundant message in instance rename when the name is given
1135
  already as a FQDN
1136
- Added option to ``gnt-instance recreate-disks`` to allow creating the
1137
  disks on new nodes, allowing recreation when the original instance
1138
  nodes are completely gone
1139
- Added option when converting disk templates to DRBD to skip waiting
1140
  for the resync, in order to make the instance available sooner
1141
- Added two new variables to the OS scripts environment (containing the
1142
  instance's nodes)
1143
- Made the root_path and optional parameter for the xen-pvm hypervisor,
1144
  to allow use of ``pvgrub`` as bootloader
1145
- Changed the instance memory modifications to only check out-of-memory
1146
  conditions on memory increases, and turned the secondary node warnings
1147
  into errors (they can still be overridden via ``--force``)
1148
- Fixed the handling of a corner case when the Python installation gets
1149
  corrupted (e.g. a bad disk) while ganeti-noded is running and we try
1150
  to execute a command that doesn't exist
1151
- Fixed a bug in ``gnt-instance move`` (LUInstanceMove) when the primary
1152
  node of the instance returned failures during instance shutdown; this
1153
  adds the option ``--ignore-consistency`` to gnt-instance move
1154

    
1155
And as usual, various improvements to the error messages, documentation
1156
and man pages.
1157

    
1158

    
1159
Version 2.4.1
1160
-------------
1161

    
1162
*(Released Wed, 09 Mar 2011)*
1163

    
1164
Emergency bug-fix release. ``tools/cfgupgrade`` was broken and overwrote
1165
the RAPI users file if run twice (even with ``--dry-run``).
1166

    
1167
The release fixes that bug (nothing else changed).
1168

    
1169

    
1170
Version 2.4.0
1171
-------------
1172

    
1173
*(Released Mon, 07 Mar 2011)*
1174

    
1175
Final 2.4.0 release. Just a few small fixes:
1176

    
1177
- Fixed RAPI node evacuate
1178
- Fixed the kvm-ifup script
1179
- Fixed internal error handling for special job cases
1180
- Updated man page to specify the escaping feature for options
1181

    
1182

    
1183
Version 2.4.0 rc3
1184
-----------------
1185

    
1186
*(Released Mon, 28 Feb 2011)*
1187

    
1188
A critical fix for the ``prealloc_wipe_disks`` feature: it is possible
1189
that this feature wiped the disks of the wrong instance, leading to loss
1190
of data.
1191

    
1192
Other changes:
1193

    
1194
- Fixed title of query field containing instance name
1195
- Expanded the glossary in the documentation
1196
- Fixed one unittest (internal issue)
1197

    
1198

    
1199
Version 2.4.0 rc2
1200
-----------------
1201

    
1202
*(Released Mon, 21 Feb 2011)*
1203

    
1204
A number of bug fixes plus just a couple functionality changes.
1205

    
1206
On the user-visible side, the ``gnt-* list`` command output has changed
1207
with respect to "special" field states. The current rc1 style of display
1208
can be re-enabled by passing a new ``--verbose`` (``-v``) flag, but in
1209
the default output mode special fields states are displayed as follows:
1210

    
1211
- Offline resource: ``*``
1212
- Unavailable/not applicable: ``-``
1213
- Data missing (RPC failure): ``?``
1214
- Unknown field: ``??``
1215

    
1216
Another user-visible change is the addition of ``--force-join`` to
1217
``gnt-node add``.
1218

    
1219
As for bug fixes:
1220

    
1221
- ``tools/cluster-merge`` has seen many fixes and is now enabled again
1222
- Fixed regression in RAPI/instance reinstall where all parameters were
1223
  required (instead of optional)
1224
- Fixed ``gnt-cluster repair-disk-sizes``, was broken since Ganeti 2.2
1225
- Fixed iallocator usage (offline nodes were not considered offline)
1226
- Fixed ``gnt-node list`` with respect to non-vm_capable nodes
1227
- Fixed hypervisor and OS parameter validation with respect to
1228
  non-vm_capable nodes
1229
- Fixed ``gnt-cluster verify`` with respect to offline nodes (mostly
1230
  cosmetic)
1231
- Fixed ``tools/listrunner`` with respect to agent-based usage
1232

    
1233

    
1234
Version 2.4.0 rc1
1235
-----------------
1236

    
1237
*(Released Fri,  4 Feb 2011)*
1238

    
1239
Many changes and fixes since the beta1 release. While there were some
1240
internal changes, the code has been mostly stabilised for the RC
1241
release.
1242

    
1243
Note: the dumb allocator was removed in this release, as it was not kept
1244
up-to-date with the IAllocator protocol changes. It is recommended to
1245
use the ``hail`` command from the ganeti-htools package.
1246

    
1247
Note: the 2.4 and up versions of Ganeti are not compatible with the
1248
0.2.x branch of ganeti-htools. You need to upgrade to
1249
ganeti-htools-0.3.0 (or later).
1250

    
1251
Regressions fixed from 2.3
1252
~~~~~~~~~~~~~~~~~~~~~~~~~~
1253

    
1254
- Fixed the ``gnt-cluster verify-disks`` command
1255
- Made ``gnt-cluster verify-disks`` work in parallel (as opposed to
1256
  serially on nodes)
1257
- Fixed disk adoption breakage
1258
- Fixed wrong headers in instance listing for field aliases
1259

    
1260
Other bugs fixed
1261
~~~~~~~~~~~~~~~~
1262

    
1263
- Fixed corner case in KVM handling of NICs
1264
- Fixed many cases of wrong handling of non-vm_capable nodes
1265
- Fixed a bug where a missing instance symlink was not possible to
1266
  recreate with any ``gnt-*`` command (now ``gnt-instance
1267
  activate-disks`` does it)
1268
- Fixed the volume group name as reported by ``gnt-cluster
1269
  verify-disks``
1270
- Increased timeouts for the import-export code, hopefully leading to
1271
  fewer aborts due network or instance timeouts
1272
- Fixed bug in ``gnt-node list-storage``
1273
- Fixed bug where not all daemons were started on cluster
1274
  initialisation, but only at the first watcher run
1275
- Fixed many bugs in the OOB implementation
1276
- Fixed watcher behaviour in presence of instances with offline
1277
  secondaries
1278
- Fixed instance list output for instances running on the wrong node
1279
- a few fixes to the cluster-merge tool, but it still cannot merge
1280
  multi-node groups (currently it is not recommended to use this tool)
1281

    
1282

    
1283
Improvements
1284
~~~~~~~~~~~~
1285

    
1286
- Improved network configuration for the KVM hypervisor
1287
- Added e1000 as a supported NIC for Xen-HVM
1288
- Improved the lvmstrap tool to also be able to use partitions, as
1289
  opposed to full disks
1290
- Improved speed of disk wiping (the cluster parameter
1291
  ``prealloc_wipe_disks``, so that it has a low impact on the total time
1292
  of instance creations
1293
- Added documentation for the OS parameters
1294
- Changed ``gnt-instance deactivate-disks`` so that it can work if the
1295
  hypervisor is not responding
1296
- Added display of blacklisted and hidden OS information in
1297
  ``gnt-cluster info``
1298
- Extended ``gnt-cluster verify`` to also validate hypervisor, backend,
1299
  NIC and node parameters, which might create problems with currently
1300
  invalid (but undetected) configuration files, but prevents validation
1301
  failures when unrelated parameters are modified
1302
- Changed cluster initialisation to wait for the master daemon to become
1303
  available
1304
- Expanded the RAPI interface:
1305

    
1306
  - Added config redistribution resource
1307
  - Added activation/deactivation of instance disks
1308
  - Added export of console information
1309

    
1310
- Implemented log file reopening on SIGHUP, which allows using
1311
  logrotate(8) for the Ganeti log files
1312
- Added a basic OOB helper script as an example
1313

    
1314

    
1315
Version 2.4.0 beta1
1316
-------------------
1317

    
1318
*(Released Fri, 14 Jan 2011)*
1319

    
1320
User-visible
1321
~~~~~~~~~~~~
1322

    
1323
- Fixed timezone issues when formatting timestamps
1324
- Added support for node groups, available via ``gnt-group`` and other
1325
  commands
1326
- Added out-of-band framework and management, see :doc:`design
1327
  document <design-oob>`
1328
- Removed support for roman numbers from ``gnt-node list`` and
1329
  ``gnt-instance list``.
1330
- Allowed modification of master network interface via ``gnt-cluster
1331
  modify --master-netdev``
1332
- Accept offline secondaries while shutting down instance disks
1333
- Added ``blockdev_prefix`` parameter to Xen PVM and HVM hypervisors
1334
- Added support for multiple LVM volume groups
1335
- Avoid sorting nodes for ``gnt-node list`` if specific nodes are
1336
  requested
1337
- Added commands to list available fields:
1338

    
1339
  - ``gnt-node list-fields``
1340
  - ``gnt-group list-fields``
1341
  - ``gnt-instance list-fields``
1342

    
1343
- Updated documentation and man pages
1344

    
1345
Integration
1346
~~~~~~~~~~~
1347

    
1348
- Moved ``rapi_users`` file into separate directory, now named
1349
  ``.../ganeti/rapi/users``, ``cfgupgrade`` moves the file and creates a
1350
  symlink
1351
- Added new tool for running commands on many machines,
1352
  ``tools/ganeti-listrunner``
1353
- Implemented more verbose result in ``OpInstanceConsole`` opcode, also
1354
  improving the ``gnt-instance console`` output
1355
- Allowed customisation of disk index separator at ``configure`` time
1356
- Export node group allocation policy to :doc:`iallocator <iallocator>`
1357
- Added support for non-partitioned md disks in ``lvmstrap``
1358
- Added script to gracefully power off KVM instances
1359
- Split ``utils`` module into smaller parts
1360
- Changed query operations to return more detailed information, e.g.
1361
  whether an information is unavailable due to an offline node. To use
1362
  this new functionality, the LUXI call ``Query`` must be used. Field
1363
  information is now stored by the master daemon and can be retrieved
1364
  using ``QueryFields``. Instances, nodes and groups can also be queried
1365
  using the new opcodes ``OpQuery`` and ``OpQueryFields`` (not yet
1366
  exposed via RAPI). The following commands make use of this
1367
  infrastructure change:
1368

    
1369
  - ``gnt-group list``
1370
  - ``gnt-group list-fields``
1371
  - ``gnt-node list``
1372
  - ``gnt-node list-fields``
1373
  - ``gnt-instance list``
1374
  - ``gnt-instance list-fields``
1375
  - ``gnt-debug locks``
1376

    
1377
Remote API
1378
~~~~~~~~~~
1379

    
1380
- New RAPI resources (see :doc:`rapi`):
1381

    
1382
  - ``/2/modify``
1383
  - ``/2/groups``
1384
  - ``/2/groups/[group_name]``
1385
  - ``/2/groups/[group_name]/assign-nodes``
1386
  - ``/2/groups/[group_name]/modify``
1387
  - ``/2/groups/[group_name]/rename``
1388
  - ``/2/instances/[instance_name]/disk/[disk_index]/grow``
1389

    
1390
- RAPI changes:
1391

    
1392
  - Implemented ``no_install`` for instance creation
1393
  - Implemented OS parameters for instance reinstallation, allowing
1394
    use of special settings on reinstallation (e.g. for preserving data)
1395

    
1396
Misc
1397
~~~~
1398

    
1399
- Added IPv6 support in import/export
1400
- Pause DRBD synchronization while wiping disks on instance creation
1401
- Updated unittests and QA scripts
1402
- Improved network parameters passed to KVM
1403
- Converted man pages from docbook to reStructuredText
1404

    
1405

    
1406
Version 2.3.1
1407
-------------
1408

    
1409
*(Released Mon, 20 Dec 2010)*
1410

    
1411
Released version 2.3.1~rc1 without any changes.
1412

    
1413

    
1414
Version 2.3.1 rc1
1415
-----------------
1416

    
1417
*(Released Wed, 1 Dec 2010)*
1418

    
1419
- impexpd: Disable OpenSSL compression in socat if possible (backport
1420
  from master, commit e90739d625b, see :doc:`installation guide
1421
  <install-quick>` for details)
1422
- Changed unittest coverage report to exclude test scripts
1423
- Added script to check version format
1424

    
1425

    
1426
Version 2.3.0
1427
-------------
1428

    
1429
*(Released Wed, 1 Dec 2010)*
1430

    
1431
Released version 2.3.0~rc1 without any changes.
1432

    
1433

    
1434
Version 2.3.0 rc1
1435
-----------------
1436

    
1437
*(Released Fri, 19 Nov 2010)*
1438

    
1439
A number of bugfixes and documentation updates:
1440

    
1441
- Update ganeti-os-interface documentation
1442
- Fixed a bug related to duplicate MACs or similar items which should be
1443
  unique
1444
- Fix breakage in OS state modify
1445
- Reinstall instance: disallow offline secondaries (fixes bug related to
1446
  OS changing but reinstall failing)
1447
- plus all the other fixes between 2.2.1 and 2.2.2
1448

    
1449

    
1450
Version 2.3.0 rc0
1451
-----------------
1452

    
1453
*(Released Tue, 2 Nov 2010)*
1454

    
1455
- Fixed clearing of the default iallocator using ``gnt-cluster modify``
1456
- Fixed master failover race with watcher
1457
- Fixed a bug in ``gnt-node modify`` which could lead to an inconsistent
1458
  configuration
1459
- Accept previously stopped instance for export with instance removal
1460
- Simplify and extend the environment variables for instance OS scripts
1461
- Added new node flags, ``master_capable`` and ``vm_capable``
1462
- Added optional instance disk wiping prior during allocation. This is a
1463
  cluster-wide option and can be set/modified using
1464
  ``gnt-cluster {init,modify} --prealloc-wipe-disks``.
1465
- Added IPv6 support, see :doc:`design document <design-2.3>` and
1466
  :doc:`install-quick`
1467
- Added a new watcher option (``--ignore-pause``)
1468
- Added option to ignore offline node on instance start/stop
1469
  (``--ignore-offline``)
1470
- Allow overriding OS parameters with ``gnt-instance reinstall``
1471
- Added ability to change node's secondary IP address using ``gnt-node
1472
  modify``
1473
- Implemented privilege separation for all daemons except
1474
  ``ganeti-noded``, see ``configure`` options
1475
- Complain if an instance's disk is marked faulty in ``gnt-cluster
1476
  verify``
1477
- Implemented job priorities (see ``ganeti(7)`` manpage)
1478
- Ignore failures while shutting down instances during failover from
1479
  offline node
1480
- Exit daemon's bootstrap process only once daemon is ready
1481
- Export more information via ``LUInstanceQuery``/remote API
1482
- Improved documentation, QA and unittests
1483
- RAPI daemon now watches ``rapi_users`` all the time and doesn't need a
1484
  restart if the file was created or changed
1485
- Added LUXI protocol version sent with each request and response,
1486
  allowing detection of server/client mismatches
1487
- Moved the Python scripts among gnt-* and ganeti-* into modules
1488
- Moved all code related to setting up SSH to an external script,
1489
  ``setup-ssh``
1490
- Infrastructure changes for node group support in future versions
1491

    
1492

    
1493
Version 2.2.2
1494
-------------
1495

    
1496
*(Released Fri, 19 Nov 2010)*
1497

    
1498
A few small bugs fixed, and some improvements to the build system:
1499

    
1500
- Fix documentation regarding conversion to drbd
1501
- Fix validation of parameters in cluster modify (``gnt-cluster modify
1502
  -B``)
1503
- Fix error handling in node modify with multiple changes
1504
- Allow remote imports without checked names
1505

    
1506

    
1507
Version 2.2.1
1508
-------------
1509

    
1510
*(Released Tue, 19 Oct 2010)*
1511

    
1512
- Disable SSL session ID cache in RPC client
1513

    
1514

    
1515
Version 2.2.1 rc1
1516
-----------------
1517

    
1518
*(Released Thu, 14 Oct 2010)*
1519

    
1520
- Fix interaction between Curl/GnuTLS and the Python's HTTP server
1521
  (thanks Apollon Oikonomopoulos!), finally allowing the use of Curl
1522
  with GnuTLS
1523
- Fix problems with interaction between Curl and Python's HTTP server,
1524
  resulting in increased speed in many RPC calls
1525
- Improve our release script to prevent breakage with older aclocal and
1526
  Python 2.6
1527

    
1528

    
1529
Version 2.2.1 rc0
1530
-----------------
1531

    
1532
*(Released Thu, 7 Oct 2010)*
1533

    
1534
- Fixed issue 125, replace hardcoded "xenvg" in ``gnt-cluster`` with
1535
  value retrieved from master
1536
- Added support for blacklisted or hidden OS definitions
1537
- Added simple lock monitor (accessible via (``gnt-debug locks``)
1538
- Added support for -mem-path in KVM hypervisor abstraction layer
1539
- Allow overriding instance parameters in tool for inter-cluster
1540
  instance moves (``tools/move-instance``)
1541
- Improved opcode summaries (e.g. in ``gnt-job list``)
1542
- Improve consistency of OS listing by sorting it
1543
- Documentation updates
1544

    
1545

    
1546
Version 2.2.0.1
1547
---------------
1548

    
1549
*(Released Fri, 8 Oct 2010)*
1550

    
1551
- Rebuild with a newer autotools version, to fix python 2.6 compatibility
1552

    
1553

    
1554
Version 2.2.0
1555
-------------
1556

    
1557
*(Released Mon, 4 Oct 2010)*
1558

    
1559
- Fixed regression in ``gnt-instance rename``
1560

    
1561

    
1562
Version 2.2.0 rc2
1563
-----------------
1564

    
1565
*(Released Wed, 22 Sep 2010)*
1566

    
1567
- Fixed OS_VARIANT variable for OS scripts
1568
- Fixed cluster tag operations via RAPI
1569
- Made ``setup-ssh`` exit with non-zero code if an error occurred
1570
- Disabled RAPI CA checks in watcher
1571

    
1572

    
1573
Version 2.2.0 rc1
1574
-----------------
1575

    
1576
*(Released Mon, 23 Aug 2010)*
1577

    
1578
- Support DRBD versions of the format "a.b.c.d"
1579
- Updated manpages
1580
- Re-introduce support for usage from multiple threads in RAPI client
1581
- Instance renames and modify via RAPI
1582
- Work around race condition between processing and archival in job
1583
  queue
1584
- Mark opcodes following failed one as failed, too
1585
- Job field ``lock_status`` was removed due to difficulties making it
1586
  work with the changed job queue in Ganeti 2.2; a better way to monitor
1587
  locks is expected for a later 2.2.x release
1588
- Fixed dry-run behaviour with many commands
1589
- Support ``ssh-agent`` again when adding nodes
1590
- Many additional bugfixes
1591

    
1592

    
1593
Version 2.2.0 rc0
1594
-----------------
1595

    
1596
*(Released Fri, 30 Jul 2010)*
1597

    
1598
Important change: the internal RPC mechanism between Ganeti nodes has
1599
changed from using a home-grown http library (based on the Python base
1600
libraries) to use the PycURL library. This requires that PycURL is
1601
installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
1602
against GnuTLS by default. cURL's support for GnuTLS had known issues
1603
before cURL 7.21.0 and we recommend using the latest cURL release or
1604
linking against OpenSSL. Most other distributions already link PycURL
1605
and cURL against OpenSSL. The command::
1606

    
1607
  python -c 'import pycurl; print pycurl.version'
1608

    
1609
can be used to determine the libraries PycURL and cURL are linked
1610
against.
1611

    
1612
Other significant changes:
1613

    
1614
- Rewrote much of the internals of the job queue, in order to achieve
1615
  better parallelism; this decouples job query operations from the job
1616
  processing, and it should allow much nicer behaviour of the master
1617
  daemon under load, and it also has uncovered some long-standing bugs
1618
  related to the job serialisation (now fixed)
1619
- Added a default iallocator setting to the cluster parameters,
1620
  eliminating the need to always pass nodes or an iallocator for
1621
  operations that require selection of new node(s)
1622
- Added experimental support for the LXC virtualization method
1623
- Added support for OS parameters, which allows the installation of
1624
  instances to pass parameter to OS scripts in order to customise the
1625
  instance
1626
- Added a hypervisor parameter controlling the migration type (live or
1627
  non-live), since hypervisors have various levels of reliability; this
1628
  has renamed the 'live' parameter to 'mode'
1629
- Added a cluster parameter ``reserved_lvs`` that denotes reserved
1630
  logical volumes, meaning that cluster verify will ignore them and not
1631
  flag their presence as errors
1632
- The watcher will now reset the error count for failed instances after
1633
  8 hours, thus allowing self-healing if the problem that caused the
1634
  instances to be down/fail to start has cleared in the meantime
1635
- Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
1636
  check for, and warn, if the drbd module parameter ``usermode_helper``
1637
  is not consistent with the cluster-wide setting; this is needed to
1638
  make diagnose easier of failed drbd creations
1639
- Started adding base IPv6 support, but this is not yet
1640
  enabled/available for use
1641
- Rename operations (cluster, instance) will now return the new name,
1642
  which is especially useful if a short name was passed in
1643
- Added support for instance migration in RAPI
1644
- Added a tool to pre-configure nodes for the SSH setup, before joining
1645
  them to the cluster; this will allow in the future a simplified model
1646
  for node joining (but not yet fully enabled in 2.2); this needs the
1647
  paramiko python library
1648
- Fixed handling of name-resolving errors
1649
- Fixed consistency of job results on the error path
1650
- Fixed master-failover race condition when executed multiple times in
1651
  sequence
1652
- Fixed many bugs related to the job queue (mostly introduced during the
1653
  2.2 development cycle, so not all are impacting 2.1)
1654
- Fixed instance migration with missing disk symlinks
1655
- Fixed handling of unknown jobs in ``gnt-job archive``
1656
- And many other small fixes/improvements
1657

    
1658
Internal changes:
1659

    
1660
- Enhanced both the unittest and the QA coverage
1661
- Switched the opcode validation to a generic model, and extended the
1662
  validation to all opcode parameters
1663
- Changed more parts of the code that write shell scripts to use the
1664
  same class for this
1665
- Switched the master daemon to use the asyncore library for the Luxi
1666
  server endpoint
1667

    
1668

    
1669
Version 2.2.0 beta0
1670
-------------------
1671

    
1672
*(Released Thu, 17 Jun 2010)*
1673

    
1674
- Added tool (``move-instance``) and infrastructure to move instances
1675
  between separate clusters (see :doc:`separate documentation
1676
  <move-instance>` and :doc:`design document <design-2.2>`)
1677
- Added per-request RPC timeout
1678
- RAPI now requires a Content-Type header for requests with a body (e.g.
1679
  ``PUT`` or ``POST``) which must be set to ``application/json`` (see
1680
  :rfc:`2616` (HTTP/1.1), section 7.2.1)
1681
- ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
1682
  reachable
1683
- Implemented initial support for running Ganeti daemons as separate
1684
  users, see configure-time flags ``--with-user-prefix`` and
1685
  ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
1686
  time)
1687
- Instances can be removed after export (``gnt-backup export
1688
  --remove-instance``)
1689
- Self-signed certificates generated by Ganeti now use a 2048 bit RSA
1690
  key (instead of 1024 bit)
1691
- Added new cluster configuration file for cluster domain secret
1692
- Import/export now use SSL instead of SSH
1693
- Added support for showing estimated time when exporting an instance,
1694
  see the ``ganeti-os-interface(7)`` manpage and look for
1695
  ``EXP_SIZE_FD``
1696

    
1697

    
1698
Version 2.1.8
1699
-------------
1700

    
1701
*(Released Tue, 16 Nov 2010)*
1702

    
1703
Some more bugfixes. Unless critical bugs occur, this will be the last
1704
2.1 release:
1705

    
1706
- Fix case of MAC special-values
1707
- Fix mac checker regex
1708
- backend: Fix typo causing "out of range" error
1709
- Add missing --units in gnt-instance list man page
1710

    
1711

    
1712
Version 2.1.7
1713
-------------
1714

    
1715
*(Released Tue, 24 Aug 2010)*
1716

    
1717
Bugfixes only:
1718
  - Don't ignore secondary node silently on non-mirrored disk templates
1719
    (issue 113)
1720
  - Fix --master-netdev arg name in gnt-cluster(8) (issue 114)
1721
  - Fix usb_mouse parameter breaking with vnc_console (issue 109)
1722
  - Properly document the usb_mouse parameter
1723
  - Fix path in ganeti-rapi(8) (issue 116)
1724
  - Adjust error message when the ganeti user's .ssh directory is
1725
    missing
1726
  - Add same-node-check when changing the disk template to drbd
1727

    
1728

    
1729
Version 2.1.6
1730
-------------
1731

    
1732
*(Released Fri, 16 Jul 2010)*
1733

    
1734
Bugfixes only:
1735
  - Add an option to only select some reboot types during qa/burnin.
1736
    (on some hypervisors consequent reboots are not supported)
1737
  - Fix infrequent race condition in master failover. Sometimes the old
1738
    master ip address would be still detected as up for a short time
1739
    after it was removed, causing failover to fail.
1740
  - Decrease mlockall warnings when the ctypes module is missing. On
1741
    Python 2.4 we support running even if no ctypes module is installed,
1742
    but we were too verbose about this issue.
1743
  - Fix building on old distributions, on which man doesn't have a
1744
    --warnings option.
1745
  - Fix RAPI not to ignore the MAC address on instance creation
1746
  - Implement the old instance creation format in the RAPI client.
1747

    
1748

    
1749
Version 2.1.5
1750
-------------
1751

    
1752
*(Released Thu, 01 Jul 2010)*
1753

    
1754
A small bugfix release:
1755
  - Fix disk adoption: broken by strict --disk option checking in 2.1.4
1756
  - Fix batch-create: broken in the whole 2.1 series due to a lookup on
1757
    a non-existing option
1758
  - Fix instance create: the --force-variant option was ignored
1759
  - Improve pylint 0.21 compatibility and warnings with Python 2.6
1760
  - Fix modify node storage with non-FQDN arguments
1761
  - Fix RAPI client to authenticate under Python 2.6 when used
1762
    for more than 5 requests needing authentication
1763
  - Fix gnt-instance modify -t (storage) giving a wrong error message
1764
    when converting a non-shutdown drbd instance to plain
1765

    
1766

    
1767
Version 2.1.4
1768
-------------
1769

    
1770
*(Released Fri, 18 Jun 2010)*
1771

    
1772
A small bugfix release:
1773

    
1774
  - Fix live migration of KVM instances started with older Ganeti
1775
    versions which had fewer hypervisor parameters
1776
  - Fix gnt-instance grow-disk on down instances
1777
  - Fix an error-reporting bug during instance migration
1778
  - Better checking of the ``--net`` and ``--disk`` values, to avoid
1779
    silently ignoring broken ones
1780
  - Fix an RPC error reporting bug affecting, for example, RAPI client
1781
    users
1782
  - Fix bug triggered by different API version os-es on different nodes
1783
  - Fix a bug in instance startup with custom hvparams: OS level
1784
    parameters would fail to be applied.
1785
  - Fix the RAPI client under Python 2.6 (but more work is needed to
1786
    make it work completely well with OpenSSL)
1787
  - Fix handling of errors when resolving names from DNS
1788

    
1789

    
1790
Version 2.1.3
1791
-------------
1792

    
1793
*(Released Thu, 3 Jun 2010)*
1794

    
1795
A medium sized development cycle. Some new features, and some
1796
fixes/small improvements/cleanups.
1797

    
1798
Significant features
1799
~~~~~~~~~~~~~~~~~~~~
1800

    
1801
The node deamon now tries to mlock itself into memory, unless the
1802
``--no-mlock`` flag is passed. It also doesn't fail if it can't write
1803
its logs, and falls back to console logging. This allows emergency
1804
features such as ``gnt-node powercycle`` to work even in the event of a
1805
broken node disk (tested offlining the disk hosting the node's
1806
filesystem and dropping its memory caches; don't try this at home)
1807

    
1808
KVM: add vhost-net acceleration support. It can be tested with a new
1809
enough version of the kernel and of qemu-kvm.
1810

    
1811
KVM: Add instance chrooting feature. If you use privilege dropping for
1812
your VMs you can also now force them to chroot to an empty directory,
1813
before starting the emulated guest.
1814

    
1815
KVM: Add maximum migration bandwith and maximum downtime tweaking
1816
support (requires a new-enough version of qemu-kvm).
1817

    
1818
Cluster verify will now warn if the master node doesn't have the master
1819
ip configured on it.
1820

    
1821
Add a new (incompatible) instance creation request format to RAPI which
1822
supports all parameters (previously only a subset was supported, and it
1823
wasn't possible to extend the old format to accomodate all the new
1824
features. The old format is still supported, and a client can check for
1825
this feature, before using it, by checking for its presence in the
1826
``features`` RAPI resource.
1827

    
1828
Now with ancient latin support. Try it passing the ``--roman`` option to
1829
``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
1830
(requires the python-roman module to be installed, in order to work).
1831

    
1832
Other changes
1833
~~~~~~~~~~~~~
1834

    
1835
As usual many internal code refactorings, documentation updates, and
1836
such. Among others:
1837

    
1838
  - Lots of improvements and cleanups to the experimental Remote API
1839
    (RAPI) client library.
1840
  - A new unit test suite for the core daemon libraries.
1841
  - A fix to creating missing directories makes sure the umask is not
1842
    applied anymore. This enforces the same directory permissions
1843
    everywhere.
1844
  - Better handling terminating daemons with ctrl+c (used when running
1845
    them in debugging mode).
1846
  - Fix a race condition in live migrating a KVM instance, when stat()
1847
    on the old proc status file returned EINVAL, which is an unexpected
1848
    value.
1849
  - Fixed manpage checking with newer man and utf-8 charachters. But now
1850
    you need the en_US.UTF-8 locale enabled to build Ganeti from git.
1851

    
1852

    
1853
Version 2.1.2.1
1854
---------------
1855

    
1856
*(Released Fri, 7 May 2010)*
1857

    
1858
Fix a bug which prevented untagged KVM instances from starting.
1859

    
1860

    
1861
Version 2.1.2
1862
-------------
1863

    
1864
*(Released Fri, 7 May 2010)*
1865

    
1866
Another release with a long development cycle, during which many
1867
different features were added.
1868

    
1869
Significant features
1870
~~~~~~~~~~~~~~~~~~~~
1871

    
1872
The KVM hypervisor now can run the individual instances as non-root, to
1873
reduce the impact of a VM being hijacked due to bugs in the
1874
hypervisor. It is possible to run all instances as a single (non-root)
1875
user, to manually specify a user for each instance, or to dynamically
1876
allocate a user out of a cluster-wide pool to each instance, with the
1877
guarantee that no two instances will run under the same user ID on any
1878
given node.
1879

    
1880
An experimental RAPI client library, that can be used standalone
1881
(without the other Ganeti libraries), is provided in the source tree as
1882
``lib/rapi/client.py``. Note this client might change its interface in
1883
the future, as we iterate on its capabilities.
1884

    
1885
A new command, ``gnt-cluster renew-crypto`` has been added to easily
1886
replace the cluster's certificates and crypto keys. This might help in
1887
case they have been compromised, or have simply expired.
1888

    
1889
A new disk option for instance creation has been added that allows one
1890
to "adopt" currently existing logical volumes, with data
1891
preservation. This should allow easier migration to Ganeti from
1892
unmanaged (or managed via other software) instances.
1893

    
1894
Another disk improvement is the possibility to convert between redundant
1895
(DRBD) and plain (LVM) disk configuration for an instance. This should
1896
allow better scalability (starting with one node and growing the
1897
cluster, or shrinking a two-node cluster to one node).
1898

    
1899
A new feature that could help with automated node failovers has been
1900
implemented: if a node sees itself as offline (by querying the master
1901
candidates), it will try to shutdown (hard) all instances and any active
1902
DRBD devices. This reduces the risk of duplicate instances if an
1903
external script automatically failovers the instances on such nodes. To
1904
enable this, the cluster parameter ``maintain_node_health`` should be
1905
enabled; in the future this option (per the name) will enable other
1906
automatic maintenance features.
1907

    
1908
Instance export/import now will reuse the original instance
1909
specifications for all parameters; that means exporting an instance,
1910
deleting it and the importing it back should give an almost identical
1911
instance. Note that the default import behaviour has changed from
1912
before, where it created only one NIC; now it recreates the original
1913
number of NICs.
1914

    
1915
Cluster verify has added a few new checks: SSL certificates validity,
1916
/etc/hosts consistency across the cluster, etc.
1917

    
1918
Other changes
1919
~~~~~~~~~~~~~
1920

    
1921
As usual, many internal changes were done, documentation fixes,
1922
etc. Among others:
1923

    
1924
- Fixed cluster initialization with disabled cluster storage (regression
1925
  introduced in 2.1.1)
1926
- File-based storage supports growing the disks
1927
- Fixed behaviour of node role changes
1928
- Fixed cluster verify for some corner cases, plus a general rewrite of
1929
  cluster verify to allow future extension with more checks
1930
- Fixed log spamming by watcher and node daemon (regression introduced
1931
  in 2.1.1)
1932
- Fixed possible validation issues when changing the list of enabled
1933
  hypervisors
1934
- Fixed cleanup of /etc/hosts during node removal
1935
- Fixed RAPI response for invalid methods
1936
- Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
1937
- Multiple small improvements to the KVM hypervisor (VNC usage, booting
1938
  from ide disks, etc.)
1939
- Allow OS changes without re-installation (to record a changed OS
1940
  outside of Ganeti, or to allow OS renames)
1941
- Allow instance creation without OS installation (useful for example if
1942
  the OS will be installed manually, or restored from a backup not in
1943
  Ganeti format)
1944
- Implemented option to make cluster ``copyfile`` use the replication
1945
  network
1946
- Added list of enabled hypervisors to ssconf (possibly useful for
1947
  external scripts)
1948
- Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
1949
  1.2 clusters
1950
- A partial form of node re-IP is possible via node readd, which now
1951
  allows changed node primary IP
1952
- Command line utilities now show an informational message if the job is
1953
  waiting for a lock
1954
- The logs of the master daemon now show the PID/UID/GID of the
1955
  connected client
1956

    
1957

    
1958
Version 2.1.1
1959
-------------
1960

    
1961
*(Released Fri, 12 Mar 2010)*
1962

    
1963
During the 2.1.0 long release candidate cycle, a lot of improvements and
1964
changes have accumulated with were released later as 2.1.1.
1965

    
1966
Major changes
1967
~~~~~~~~~~~~~
1968

    
1969
The node evacuate command (``gnt-node evacuate``) was significantly
1970
rewritten, and as such the IAllocator protocol was changed - a new
1971
request type has been added. This unfortunate change during a stable
1972
series is designed to improve performance of node evacuations; on
1973
clusters with more than about five nodes and which are well-balanced,
1974
evacuation should proceed in parallel for all instances of the node
1975
being evacuated. As such, any existing IAllocator scripts need to be
1976
updated, otherwise the above command will fail due to the unknown
1977
request. The provided "dumb" allocator has not been updated; but the
1978
ganeti-htools package supports the new protocol since version 0.2.4.
1979

    
1980
Another important change is increased validation of node and instance
1981
names. This might create problems in special cases, if invalid host
1982
names are being used.
1983

    
1984
Also, a new layer of hypervisor parameters has been added, that sits at
1985
OS level between the cluster defaults and the instance ones. This allows
1986
customisation of virtualization parameters depending on the installed
1987
OS. For example instances with OS 'X' may have a different KVM kernel
1988
(or any other parameter) than the cluster defaults. This is intended to
1989
help managing a multiple OSes on the same cluster, without manual
1990
modification of each instance's parameters.
1991

    
1992
A tool for merging clusters, ``cluster-merge``, has been added in the
1993
tools sub-directory.
1994

    
1995
Bug fixes
1996
~~~~~~~~~
1997

    
1998
- Improved the int/float conversions that should make the code more
1999
  robust in face of errors from the node daemons
2000
- Fixed the remove node code in case of internal configuration errors
2001
- Fixed the node daemon behaviour in face of inconsistent queue
2002
  directory (e.g. read-only file-system where we can't open the files
2003
  read-write, etc.)
2004
- Fixed the behaviour of gnt-node modify for master candidate demotion;
2005
  now it either aborts cleanly or, if given the new "auto_promote"
2006
  parameter, will automatically promote other nodes as needed
2007
- Fixed compatibility with (unreleased yet) Python 2.6.5 that would
2008
  completely prevent Ganeti from working
2009
- Fixed bug for instance export when not all disks were successfully
2010
  exported
2011
- Fixed behaviour of node add when the new node is slow in starting up
2012
  the node daemon
2013
- Fixed handling of signals in the LUXI client, which should improve
2014
  behaviour of command-line scripts
2015
- Added checks for invalid node/instance names in the configuration (now
2016
  flagged during cluster verify)
2017
- Fixed watcher behaviour for disk activation errors
2018
- Fixed two potentially endless loops in http library, which led to the
2019
  RAPI daemon hanging and consuming 100% CPU in some cases
2020
- Fixed bug in RAPI daemon related to hashed passwords
2021
- Fixed bug for unintended qemu-level bridging of multi-NIC KVM
2022
  instances
2023
- Enhanced compatibility with non-Debian OSes, but not using absolute
2024
  path in some commands and allowing customisation of the ssh
2025
  configuration directory
2026
- Fixed possible future issue with new Python versions by abiding to the
2027
  proper use of ``__slots__`` attribute on classes
2028
- Added checks that should prevent directory traversal attacks
2029
- Many documentation fixes based on feedback from users
2030

    
2031
New features
2032
~~~~~~~~~~~~
2033

    
2034
- Added an "early_release" more for instance replace disks and node
2035
  evacuate, where we release locks earlier and thus allow higher
2036
  parallelism within the cluster
2037
- Added watcher hooks, intended to allow the watcher to restart other
2038
  daemons (e.g. from the ganeti-nbma project), but they can be used of
2039
  course for any other purpose
2040
- Added a compile-time disable for DRBD barriers, to increase
2041
  performance if the administrator trusts the power supply or the
2042
  storage system to not lose writes
2043
- Added the option of using syslog for logging instead of, or in
2044
  addition to, Ganeti's own log files
2045
- Removed boot restriction for paravirtual NICs for KVM, recent versions
2046
  can indeed boot from a paravirtual NIC
2047
- Added a generic debug level for many operations; while this is not
2048
  used widely yet, it allows one to pass the debug value all the way to
2049
  the OS scripts
2050
- Enhanced the hooks environment for instance moves (failovers,
2051
  migrations) where the primary/secondary nodes changed during the
2052
  operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
2053
- Enhanced data validations for many user-supplied values; one important
2054
  item is the restrictions imposed on instance and node names, which
2055
  might reject some (invalid) host names
2056
- Add a configure-time option to disable file-based storage, if it's not
2057
  needed; this allows greater security separation between the master
2058
  node and the other nodes from the point of view of the inter-node RPC
2059
  protocol
2060
- Added user notification in interactive tools if job is waiting in the
2061
  job queue or trying to acquire locks
2062
- Added log messages when a job is waiting for locks
2063
- Added filtering by node tags in instance operations which admit
2064
  multiple instances (start, stop, reboot, reinstall)
2065
- Added a new tool for cluster mergers, ``cluster-merge``
2066
- Parameters from command line which are of the form ``a=b,c=d`` can now
2067
  use backslash escapes to pass in values which contain commas,
2068
  e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
2069
  ``b,c``
2070
- For KVM, the instance name is the first parameter passed to KVM, so
2071
  that it's more visible in the process list
2072

    
2073

    
2074
Version 2.1.0
2075
-------------
2076

    
2077
*(Released Tue, 2 Mar 2010)*
2078

    
2079
Ganeti 2.1 brings many improvements with it. Major changes:
2080

    
2081
- Added infrastructure to ease automated disk repairs
2082
- Added new daemon to export configuration data in a cheaper way than
2083
  using the remote API
2084
- Instance NICs can now be routed instead of being associated with a
2085
  networking bridge
2086
- Improved job locking logic to reduce impact of jobs acquiring multiple
2087
  locks waiting for other long-running jobs
2088

    
2089
In-depth implementation details can be found in the Ganeti 2.1 design
2090
document.
2091

    
2092
Details
2093
~~~~~~~
2094

    
2095
- Added chroot hypervisor
2096
- Added more options to xen-hvm hypervisor (``kernel_path`` and
2097
  ``device_model``)
2098
- Added more options to xen-pvm hypervisor (``use_bootloader``,
2099
  ``bootloader_path`` and ``bootloader_args``)
2100
- Added the ``use_localtime`` option for the xen-hvm and kvm
2101
  hypervisors, and the default value for this has changed to false (in
2102
  2.0 xen-hvm always enabled it)
2103
- Added luxi call to submit multiple jobs in one go
2104
- Added cluster initialization option to not modify ``/etc/hosts``
2105
  file on nodes
2106
- Added network interface parameters
2107
- Added dry run mode to some LUs
2108
- Added RAPI resources:
2109

    
2110
  - ``/2/instances/[instance_name]/info``
2111
  - ``/2/instances/[instance_name]/replace-disks``
2112
  - ``/2/nodes/[node_name]/evacuate``
2113
  - ``/2/nodes/[node_name]/migrate``
2114
  - ``/2/nodes/[node_name]/role``
2115
  - ``/2/nodes/[node_name]/storage``
2116
  - ``/2/nodes/[node_name]/storage/modify``
2117
  - ``/2/nodes/[node_name]/storage/repair``
2118

    
2119
- Added OpCodes to evacuate or migrate all instances on a node
2120
- Added new command to list storage elements on nodes (``gnt-node
2121
  list-storage``) and modify them (``gnt-node modify-storage``)
2122
- Added new ssconf files with master candidate IP address
2123
  (``ssconf_master_candidates_ips``), node primary IP address
2124
  (``ssconf_node_primary_ips``) and node secondary IP address
2125
  (``ssconf_node_secondary_ips``)
2126
- Added ``ganeti-confd`` and a client library to query the Ganeti
2127
  configuration via UDP
2128
- Added ability to run hooks after cluster initialization and before
2129
  cluster destruction
2130
- Added automatic mode for disk replace (``gnt-instance replace-disks
2131
  --auto``)
2132
- Added ``gnt-instance recreate-disks`` to re-create (empty) disks
2133
  after catastrophic data-loss
2134
- Added ``gnt-node repair-storage`` command to repair damaged LVM volume
2135
  groups
2136
- Added ``gnt-instance move`` command to move instances
2137
- Added ``gnt-cluster watcher`` command to control watcher
2138
- Added ``gnt-node powercycle`` command to powercycle nodes
2139
- Added new job status field ``lock_status``
2140
- Added parseable error codes to cluster verification (``gnt-cluster
2141
  verify --error-codes``) and made output less verbose (use
2142
  ``--verbose`` to restore previous behaviour)
2143
- Added UUIDs to the main config entities (cluster, nodes, instances)
2144
- Added support for OS variants
2145
- Added support for hashed passwords in the Ganeti remote API users file
2146
  (``rapi_users``)
2147
- Added option to specify maximum timeout on instance shutdown
2148
- Added ``--no-ssh-init`` option to ``gnt-cluster init``
2149
- Added new helper script to start and stop Ganeti daemons
2150
  (``daemon-util``), with the intent to reduce the work necessary to
2151
  adjust Ganeti for non-Debian distributions and to start/stop daemons
2152
  from one place
2153
- Added more unittests
2154
- Fixed critical bug in ganeti-masterd startup
2155
- Removed the configure-time ``kvm-migration-port`` parameter, this is
2156
  now customisable at the cluster level for both the KVM and Xen
2157
  hypervisors using the new ``migration_port`` parameter
2158
- Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
2159
  reinstalling an instance
2160
- Allowed ``@`` in tag names
2161
- Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
2162
- Many documentation updates
2163
- Distribute hypervisor files on ``gnt-cluster redist-conf``
2164
- ``gnt-instance reinstall`` can now reinstall multiple instances
2165
- Updated many command line parameters
2166
- Introduced new OS API version 15
2167
- No longer support a default hypervisor
2168
- Treat virtual LVs as inexistent
2169
- Improved job locking logic to reduce lock contention
2170
- Match instance and node names case insensitively
2171
- Reimplemented bash completion script to be more complete
2172
- Improved burnin
2173

    
2174

    
2175
Version 2.0.6
2176
-------------
2177

    
2178
*(Released Thu, 4 Feb 2010)*
2179

    
2180
- Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
2181
- Fix a string formatting bug
2182
- Improve safety of the code in some error paths
2183
- Improve data validation in the master of values returned from nodes
2184

    
2185

    
2186
Version 2.0.5
2187
-------------
2188

    
2189
*(Released Thu, 17 Dec 2009)*
2190

    
2191
- Fix security issue due to missing validation of iallocator names; this
2192
  allows local and remote execution of arbitrary executables
2193
- Fix failure of gnt-node list during instance removal
2194
- Ship the RAPI documentation in the archive
2195

    
2196

    
2197
Version 2.0.4
2198
-------------
2199

    
2200
*(Released Wed, 30 Sep 2009)*
2201

    
2202
- Fixed many wrong messages
2203
- Fixed a few bugs related to the locking library
2204
- Fixed MAC checking at instance creation time
2205
- Fixed a DRBD parsing bug related to gaps in /proc/drbd
2206
- Fixed a few issues related to signal handling in both daemons and
2207
  scripts
2208
- Fixed the example startup script provided
2209
- Fixed insserv dependencies in the example startup script (patch from
2210
  Debian)
2211
- Fixed handling of drained nodes in the iallocator framework
2212
- Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
2213
  #528618)
2214
- Fixed error related to invalid job IDs in job polling
2215
- Fixed job/opcode persistence on unclean master shutdown
2216
- Fixed handling of partial job processing after unclean master
2217
  shutdown
2218
- Fixed error reporting from LUs, previously all errors were converted
2219
  into execution errors
2220
- Fixed error reporting from burnin
2221
- Decreased significantly the memory usage of the job queue
2222
- Optimised slightly multi-job submission
2223
- Optimised slightly opcode loading
2224
- Backported the multi-job submit framework from the development
2225
  branch; multi-instance start and stop should be faster
2226
- Added script to clean archived jobs after 21 days; this will reduce
2227
  the size of the queue directory
2228
- Added some extra checks in disk size tracking
2229
- Added an example ethers hook script
2230
- Added a cluster parameter that prevents Ganeti from modifying of
2231
  /etc/hosts
2232
- Added more node information to RAPI responses
2233
- Added a ``gnt-job watch`` command that allows following the ouput of a
2234
  job
2235
- Added a bind-address option to ganeti-rapi
2236
- Added more checks to the configuration verify
2237
- Enhanced the burnin script such that some operations can be retried
2238
  automatically
2239
- Converted instance reinstall to multi-instance model
2240

    
2241

    
2242
Version 2.0.3
2243
-------------
2244

    
2245
*(Released Fri, 7 Aug 2009)*
2246

    
2247
- Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
2248
  to allow using the pre-2.0.2 behaviour in activation, if any existing
2249
  instances have mismatched disk sizes in the configuration
2250
- Added ``gnt-cluster repair-disk-sizes`` command to check and update
2251
  any configuration mismatches for disk sizes
2252
- Added ``gnt-master cluste-failover --no-voting`` to allow master
2253
  failover to work on two-node clusters
2254
- Fixed the ``--net`` option of ``gnt-backup import``, which was
2255
  unusable
2256
- Fixed detection of OS script errors in ``gnt-backup export``
2257
- Fixed exit code of ``gnt-backup export``
2258

    
2259

    
2260
Version 2.0.2
2261
-------------
2262

    
2263
*(Released Fri, 17 Jul 2009)*
2264

    
2265
- Added experimental support for stripped logical volumes; this should
2266
  enhance performance but comes with a higher complexity in the block
2267
  device handling; stripping is only enabled when passing
2268
  ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
2269
  affected even in the non-stripped mode
2270
- Improved resiliency against transient failures at the end of DRBD
2271
  resyncs, and in general of DRBD resync checks
2272
- Fixed a couple of issues with exports and snapshot errors
2273
- Fixed a couple of issues in instance listing
2274
- Added display of the disk size in ``gnt-instance info``
2275
- Fixed checking for valid OSes in instance creation
2276
- Fixed handling of the "vcpus" parameter in instance listing and in
2277
  general of invalid parameters
2278
- Fixed http server library, and thus RAPI, to handle invalid
2279
  username/password combinations correctly; this means that now they
2280
  report unauthorized for queries too, not only for modifications,
2281
  allowing earlier detect of configuration problems
2282
- Added a new "role" node list field, equivalent to the master/master
2283
  candidate/drained/offline flags combinations
2284
- Fixed cluster modify and changes of candidate pool size
2285
- Fixed cluster verify error messages for wrong files on regular nodes
2286
- Fixed a couple of issues with node demotion from master candidate role
2287
- Fixed node readd issues
2288
- Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
2289
- Added a new ``--no-voting`` option for masterfailover to fix failover
2290
  on two-nodes clusters when the former master node is unreachable
2291
- Added instance reinstall over RAPI
2292

    
2293

    
2294
Version 2.0.1
2295
-------------
2296

    
2297
*(Released Tue, 16 Jun 2009)*
2298

    
2299
- added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
2300
  allow re-adding the start in single-user option (regression from 1.2)
2301
- the watcher writes the instance status to a file, to allow monitoring
2302
  to report the instance status (from the master) based on cached
2303
  results of the watcher's queries; while this can get stale if the
2304
  watcher is being locked due to other work on the cluster, this is
2305
  still an improvement
2306
- the watcher now also restarts the node daemon and the rapi daemon if
2307
  they died
2308
- fixed the watcher to handle full and drained queue cases
2309
- hooks export more instance data in the environment, which helps if
2310
  hook scripts need to take action based on the instance's properties
2311
  (no longer need to query back into ganeti)
2312
- instance failovers when the instance is stopped do not check for free
2313
  RAM, so that failing over a stopped instance is possible in low memory
2314
  situations
2315
- rapi uses queries for tags instead of jobs (for less job traffic), and
2316
  for cluster tags it won't talk to masterd at all but read them from
2317
  ssconf
2318
- a couple of error handling fixes in RAPI
2319
- drbd handling: improved the error handling of inconsistent disks after
2320
  resync to reduce the frequency of "there are some degraded disks for
2321
  this instance" messages
2322
- fixed a bug in live migration when DRBD doesn't want to reconnect (the
2323
  error handling path called a wrong function name)
2324

    
2325

    
2326
Version 2.0.0
2327
-------------
2328

    
2329
*(Released Wed, 27 May 2009)*
2330

    
2331
- no changes from rc5
2332

    
2333

    
2334
Version 2.0 rc5
2335
---------------
2336

    
2337
*(Released Wed, 20 May 2009)*
2338

    
2339
- fix a couple of bugs (validation, argument checks)
2340
- fix ``gnt-cluster getmaster`` on non-master nodes (regression)
2341
- some small improvements to RAPI and IAllocator
2342
- make watcher automatically start the master daemon if down
2343

    
2344

    
2345
Version 2.0 rc4
2346
---------------
2347

    
2348
*(Released Mon, 27 Apr 2009)*
2349

    
2350
- change the OS list to not require locks; this helps with big clusters
2351
- fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
2352
  volume group is broken
2353
- ``gnt-instance info``, without any arguments, doesn't run for all
2354
  instances anymore; either pass ``--all`` or pass the desired
2355
  instances; this helps against mistakes on big clusters where listing
2356
  the information for all instances takes a long time
2357
- miscellaneous doc and man pages fixes
2358

    
2359

    
2360
Version 2.0 rc3
2361
---------------
2362

    
2363
*(Released Wed, 8 Apr 2009)*
2364

    
2365
- Change the internal locking model of some ``gnt-node`` commands, in
2366
  order to reduce contention (and blocking of master daemon) when
2367
  batching many creation/reinstall jobs
2368
- Fixes to Xen soft reboot
2369
- No longer build documentation at build time, instead distribute it in
2370
  the archive, in order to reduce the need for the whole docbook/rst
2371
  toolchains
2372

    
2373

    
2374
Version 2.0 rc2
2375
---------------
2376

    
2377
*(Released Fri, 27 Mar 2009)*
2378

    
2379
- Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
2380
- Fix watcher startup sequence, improves the behaviour of busy clusters
2381
- Some other fixes in ``gnt-cluster verify``, ``gnt-instance
2382
  replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
2383
  bind address and other places
2384
- Some documentation fixes and updates
2385

    
2386

    
2387
Version 2.0 rc1
2388
---------------
2389

    
2390
*(Released Mon, 2 Mar 2009)*
2391

    
2392
- More documentation updates, now all docs should be more-or-less
2393
  up-to-date
2394
- A couple of small fixes (mixed hypervisor clusters, offline nodes,
2395
  etc.)
2396
- Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
2397
  and KVM)
2398
- Fix an issue related to $libdir/run/ganeti and cluster creation
2399

    
2400

    
2401
Version 2.0 beta2
2402
-----------------
2403

    
2404
*(Released Thu, 19 Feb 2009)*
2405

    
2406
- Xen PVM and KVM have switched the default value for the instance root
2407
  disk to the first partition on the first drive, instead of the whole
2408
  drive; this means that the OS installation scripts must be changed
2409
  accordingly
2410
- Man pages have been updated
2411
- RAPI has been switched by default to HTTPS, and the exported functions
2412
  should all work correctly
2413
- RAPI v1 has been removed
2414
- Many improvements to the KVM hypervisor
2415
- Block device errors are now better reported
2416
- Many other bugfixes and small improvements
2417

    
2418

    
2419
Version 2.0 beta1
2420
-----------------
2421

    
2422
*(Released Mon, 26 Jan 2009)*
2423

    
2424
- Version 2 is a general rewrite of the code and therefore the
2425
  differences are too many to list, see the design document for 2.0 in
2426
  the ``doc/`` subdirectory for more details
2427
- In this beta version there is not yet a migration path from 1.2 (there
2428
  will be one in the final 2.0 release)
2429
- A few significant changes are:
2430

    
2431
  - all commands are executed by a daemon (``ganeti-masterd``) and the
2432
    various ``gnt-*`` commands are just front-ends to it
2433
  - all the commands are entered into, and executed from a job queue,
2434
    see the ``gnt-job(8)`` manpage
2435
  - the RAPI daemon supports read-write operations, secured by basic
2436
    HTTP authentication on top of HTTPS
2437
  - DRBD version 0.7 support has been removed, DRBD 8 is the only
2438
    supported version (when migrating from Ganeti 1.2 to 2.0, you need
2439
    to migrate to DRBD 8 first while still running Ganeti 1.2)
2440
  - DRBD devices are using statically allocated minor numbers, which
2441
    will be assigned to existing instances during the migration process
2442
  - there is support for both Xen PVM and Xen HVM instances running on
2443
    the same cluster
2444
  - KVM virtualization is supported too
2445
  - file-based storage has been implemented, which means that it is
2446
    possible to run the cluster without LVM and DRBD storage, for
2447
    example using a shared filesystem exported from shared storage (and
2448
    still have live migration)
2449

    
2450

    
2451
Version 1.2.7
2452
-------------
2453

    
2454
*(Released Tue, 13 Jan 2009)*
2455

    
2456
- Change the default reboot type in ``gnt-instance reboot`` to "hard"
2457
- Reuse the old instance mac address by default on instance import, if
2458
  the instance name is the same.
2459
- Handle situations in which the node info rpc returns incomplete
2460
  results (issue 46)
2461
- Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
2462
- Improved version of batcher:
2463

    
2464
  - state file support
2465
  - instance mac address support
2466
  - support for HVM clusters/instances
2467

    
2468
- Add an option to show the number of cpu sockets and nodes in
2469
  ``gnt-node list``
2470
- Support OSes that handle more than one version of the OS api (but do
2471
  not change the current API in any other way)
2472
- Fix ``gnt-node migrate``
2473
- ``gnt-debug`` man page
2474
- Fixes various more typos and small issues
2475
- Increase disk resync maximum speed to 60MB/s (from 30MB/s)
2476

    
2477

    
2478
Version 1.2.6
2479
-------------
2480

    
2481
*(Released Wed, 24 Sep 2008)*
2482

    
2483
- new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
2484
  type of disk exported to fully virtualized instances.
2485
- provide access to the serial console of HVM instances
2486
- instance auto_balance flag, set by default. If turned off it will
2487
  avoid warnings on cluster verify if there is not enough memory to fail
2488
  over an instance. in the future it will prevent automatically failing
2489
  it over when we will support that.
2490
- batcher tool for instance creation, see ``tools/README.batcher``
2491
- ``gnt-instance reinstall --select-os`` to interactively select a new
2492
  operating system when reinstalling an instance.
2493
- when changing the memory amount on instance modify a check has been
2494
  added that the instance will be able to start. also warnings are
2495
  emitted if the instance will not be able to fail over, if auto_balance
2496
  is true.
2497
- documentation fixes
2498
- sync fields between ``gnt-instance list/modify/add/import``
2499
- fix a race condition in drbd when the sync speed was set after giving
2500
  the device a remote peer.
2501

    
2502

    
2503
Version 1.2.5
2504
-------------
2505

    
2506
*(Released Tue, 22 Jul 2008)*
2507

    
2508
- note: the allowed size and number of tags per object were reduced
2509
- fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
2510
- fixed twisted 8.x compatibility
2511
- fixed ``gnt-instance replace-disks`` with iallocator
2512
- add TCP keepalives on twisted connections to detect restarted nodes
2513
- disk increase support, see ``gnt-instance grow-disk``
2514
- implement bulk node/instance query for RAPI
2515
- add tags in node/instance listing (optional)
2516
- experimental migration (and live migration) support, read the man page
2517
  for ``gnt-instance migrate``
2518
- the ``ganeti-watcher`` logs are now timestamped, and the watcher also
2519
  has some small improvements in handling its state file
2520

    
2521

    
2522
Version 1.2.4
2523
-------------
2524

    
2525
*(Released Fri, 13 Jun 2008)*
2526

    
2527
- Experimental readonly, REST-based remote API implementation;
2528
  automatically started on master node, TCP port 5080, if enabled by
2529
  ``--enable-rapi`` parameter to configure script.
2530
- Instance allocator support. Add and import instance accept a
2531
  ``--iallocator`` parameter, and call that instance allocator to decide
2532
  which node to use for the instance. The iallocator document describes
2533
  what's expected from an allocator script.
2534
- ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
2535
  ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
2536
  node is lost there is still enough memory to fail over the instances
2537
  that reside on it.
2538
- ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
2539
  ``gnt-cluster verify``, to check for site-specific compliance. All the
2540
  hooks will run, and their output, if any, will be displayed. Any
2541
  failing hook will make the verification return an error value.
2542
- ``gnt-cluster verify`` now checks that its peers are reachable on the
2543
  primary and secondary interfaces
2544
- ``gnt-node add`` now supports the ``--readd`` option, to readd a node
2545
  that is still declared as part of the cluster and has failed.
2546
- ``gnt-* list`` commands now accept a new ``-o +field`` way of
2547
  specifying output fields, that just adds the chosen fields to the
2548
  default ones.
2549
- ``gnt-backup`` now has a new ``remove`` command to delete an existing
2550
  export from the filesystem.
2551
- New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
2552
  have been added. Using them you can enable/disable acpi and pae
2553
  support, and specify a path for a cd image to be exported to the
2554
  instance. These parameters as the name suggest only work on HVM
2555
  clusters.
2556
- When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
2557
  PAE support will be set to the previously hardcoded values, but the
2558
  (previously hardcoded) path to the CDROM ISO image will be unset and
2559
  if required, needs to be set manually with ``gnt-instance modify``
2560
  after the upgrade.
2561
- The address to which an instance's VNC console is bound is now
2562
  selectable per-instance, rather than being cluster wide. Of course
2563
  this only applies to instances controlled via VNC, so currently just
2564
  applies to HVM clusters.
2565

    
2566

    
2567
Version 1.2.3
2568
-------------
2569

    
2570
*(Released Mon, 18 Feb 2008)*
2571

    
2572
- more tweaks to the disk activation code (especially helpful for DRBD)
2573
- change the default ``gnt-instance list`` output format, now there is
2574
  one combined status field (see the manpage for the exact values this
2575
  field will have)
2576
- some more fixes for the mac export to hooks change
2577
- make Ganeti not break with DRBD 8.2.x (which changed the version
2578
  format in ``/proc/drbd``) (issue 24)
2579
- add an upgrade tool from "remote_raid1" disk template to "drbd" disk
2580
  template, allowing migration from DRBD0.7+MD to DRBD8
2581

    
2582

    
2583
Version 1.2.2
2584
-------------
2585

    
2586
*(Released Wed, 30 Jan 2008)*
2587

    
2588
- fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
2589
  support (issue 23)
2590
- add command aliases infrastructure and a few aliases
2591
- allow listing of VCPUs in the ``gnt-instance list`` and improve the
2592
  man pages and the ``--help`` option of ``gnt-node
2593
  list``/``gnt-instance list``
2594
- fix ``gnt-backup list`` with down nodes (issue 21)
2595
- change the tools location (move from $pkgdatadir to $pkglibdir/tools)
2596
- fix the dist archive and add a check for including svn/git files in
2597
  the future
2598
- some developer-related changes: improve the burnin and the QA suite,
2599
  add an upload script for testing during development
2600

    
2601

    
2602
Version 1.2.1
2603
-------------
2604

    
2605
*(Released Wed, 16 Jan 2008)*
2606

    
2607
- experimental HVM support, read the install document, section
2608
  "Initializing the cluster"
2609
- allow for the PVM hypervisor per-instance kernel and initrd paths
2610
- add a new command ``gnt-cluster verify-disks`` which uses a new
2611
  algorithm to improve the reconnection of the DRBD pairs if the device
2612
  on the secondary node has gone away
2613
- make logical volume code auto-activate LVs at disk activation time
2614
- slightly improve the speed of activating disks
2615
- allow specification of the MAC address at instance creation time, and
2616
  changing it later via ``gnt-instance modify``
2617
- fix handling of external commands that generate lots of output on
2618
  stderr
2619
- update documentation with regard to minimum version of DRBD8 supported
2620

    
2621

    
2622
Version 1.2.0
2623
-------------
2624

    
2625
*(Released Tue, 4 Dec 2007)*
2626

    
2627
- Log the ``xm create`` output to the node daemon log on failure (to
2628
  help diagnosing the error)
2629
- In debug mode, log all external commands output if failed to the logs
2630
- Change parsing of lvm commands to ignore stderr
2631

    
2632

    
2633
Version 1.2 beta3
2634
-----------------
2635

    
2636
*(Released Wed, 28 Nov 2007)*
2637

    
2638
- Another round of updates to the DRBD 8 code to deal with more failures
2639
  in the replace secondary node operation
2640
- Some more logging of failures in disk operations (lvm, drbd)
2641
- A few documentation updates
2642
- QA updates
2643

    
2644

    
2645
Version 1.2 beta2
2646
-----------------
2647

    
2648
*(Released Tue, 13 Nov 2007)*
2649

    
2650
- Change configuration file format from Python's Pickle to JSON.
2651
  Upgrading is possible using the cfgupgrade utility.
2652
- Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
2653
  faster replace disks and is more stable (DRBD 8 has many improvements
2654
  compared to DRBD 0.7)
2655
- Added command line tags support (see man pages for ``gnt-instance``,
2656
  ``gnt-node``, ``gnt-cluster``)
2657
- Added instance rename support
2658
- Added multi-instance startup/shutdown
2659
- Added cluster rename support
2660
- Added ``gnt-node evacuate`` to simplify some node operations
2661
- Added instance reboot operation that can speedup reboot as compared to
2662
  stop and start
2663
- Soften the requirement that hostnames are in FQDN format
2664
- The ``ganeti-watcher`` now activates drbd pairs after secondary node
2665
  reboots
2666
- Removed dependency on debian's patched fping that uses the
2667
  non-standard ``-S`` option
2668
- Now the OS definitions are searched for in multiple, configurable
2669
  paths (easier for distros to package)
2670
- Some changes to the hooks infrastructure (especially the new
2671
  post-configuration update hook)
2672
- Other small bugfixes
2673

    
2674
.. vim: set textwidth=72 syntax=rst :
2675
.. Local Variables:
2676
.. mode: rst
2677
.. fill-column: 72
2678
.. End: