8 Ganeti 2.1 brings many improvements with it. Major changes:
10 - Added infrastructure to ease automated disk repairs
11 - Added new daemon to export configuration data in a cheaper way than
13 - Instance NICs can now be routed instead of being associated with a
15 - Improved job locking logic to reduce impact of jobs acquiring multiple
16 locks waiting for other long-running jobs
18 Detailed implementation details can be found in the Ganeti 2.1 design
24 - Added chroot hypervisor
25 - Added more options to xen-hvm hypervisor (``kernel_path`` and
27 - Added more options to xen-pvm hypervisor (``use_bootloader``,
28 ``bootloader_path`` and ``bootloader_args``)
29 - Added luxi call to submit multiple jobs in one go
30 - Added cluster initialization time option to not modify ``/etc/hosts``
32 - Added network interface parameters
33 - Added dry run mode to some LUs
34 - Added RAPI resources:
36 - ``/2/instances/[instance_name]/info``
37 - ``/2/instances/[instance_name]/replace-disks``
38 - ``/2/nodes/[node_name]/evacuate``
39 - ``/2/nodes/[node_name]/migrate``
40 - ``/2/nodes/[node_name]/role``
41 - ``/2/nodes/[node_name]/storage``
42 - ``/2/nodes/[node_name]/storage/modify``
43 - ``/2/nodes/[node_name]/storage/repair``
45 - Added OpCodes to evacuate or migrate all instances on a node
46 - Added new command to list physical volumes on nodes (``gnt-node
47 physical-volumes``) and to modify them (``gnt-node modify-volume``)
48 - Added new ssconf files with master candidate IP address
49 (``ssconf_master_candidates_ips``), node primary IP address
50 (``ssconf_node_primary_ips``) and node secondary IP address
51 (``ssconf_node_secondary_ips``)
52 - Added ``ganeti-confd`` and a client library to query the Ganeti
54 - Added ability to run hooks after cluster initialization and before
56 - Added automatic mode for disk replace (``gnt-instance replace-disks
58 - Added ``gnt-node repair-volume`` command to repair damanged LVM volume
60 - Added ``gnt-instance move`` command to move instances
61 - Added ``gnt-cluster watcher`` command to control watcher
62 - Added ``gnt-node powercycle`` command to powercycle nodes
63 - Added new job status field ``lock_status``
64 - Added parseable error codes to cluster verification (``gnt-cluster
65 verify --error-codes``) and made output less verbose (use
66 ``--verbose`` to restore previous behaviour)
67 - Added UUIDs to all config entities
68 - Added support for OS variants
69 - Added more unittests
70 - Fixed critical bug in ganeti-masterd startup
71 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
72 reinstalling an instance
73 - Converted to Sphinx (http://sphinx.pocoo.org/) for documentation
74 - Many documentation updates
75 - Distribute hypervisor files on ``gnt-cluster redist-conf``
76 - ``gnt-instance reinstall`` can now reinstall multiple instances
77 - Updated many command line parameters
78 - Introduced new OS API version 15
79 - No longer support a default hypervisor
80 - Treat virtual LVs as inexistent
81 - Improved job locking logic to reduce lock contention
82 - Match instance and node names case insensitively
83 - Reimplemented bash completion script to be more complete
90 - Fixed many wrong messages
91 - Fixed a few bugs related to the locking library
92 - Fixed MAC checking at instance creation time
93 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
94 - Fixed a few issues related to signal handling in both daemons and
96 - Fixed the example startup script provided
97 - Fixed insserv dependencies in the example startup script (patch from
99 - Fixed handling of drained nodes in the iallocator framework
100 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
102 - Fixed error related to invalid job IDs in job polling
103 - Fixed job/opcode persistence on unclean master shutdown
104 - Fixed handling of partial job processing after unclean master
106 - Fixed error reporting from LUs, previously all errors were converted
107 into execution errors
108 - Fixed error reporting from burnin
109 - Decreased significantly the memory usage of the job queue
110 - Optimised slightly multi-job submission
111 - Optimised slightly opcode loading
112 - Backported the multi-job submit framework from the development
113 branch; multi-instance start and stop should be faster
114 - Added script to clean archived jobs after 21 days; this will reduce
115 the size of the queue directory
116 - Added some extra checks in disk size tracking
117 - Added an example ethers hook script
118 - Added a cluster parameter that prevents Ganeti from modifying of
120 - Added more node information to RAPI responses
121 - Added a ``gnt-job watch`` command that allows following the ouput of a
123 - Added a bind-address option to ganeti-rapi
124 - Added more checks to the configuration verify
125 - Enhanced the burnin script such that some operations can be retried
127 - Converted instance reinstall to multi-instance model
133 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
134 to allow using the pre-2.0.2 behaviour in activation, if any existing
135 instances have mismatched disk sizes in the configuration
136 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
137 any configuration mismatches for disk sizes
138 - Added ``gnt-master cluste-failover --no-voting`` to allow master
139 failover to work on two-node clusters
140 - Fixed the ``--net`` option of ``gnt-backup import``, which was
142 - Fixed detection of OS script errors in ``gnt-backup export``
143 - Fixed exit code of ``gnt-backup export``
149 - Added experimental support for stripped logical volumes; this should
150 enhance performance but comes with a higher complexity in the block
151 device handling; stripping is only enabled when passing
152 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
153 affected even in the non-stripped mode
154 - Improved resiliency against transient failures at the end of DRBD
155 resyncs, and in general of DRBD resync checks
156 - Fixed a couple of issues with exports and snapshot errors
157 - Fixed a couple of issues in instance listing
158 - Added display of the disk size in ``gnt-instance info``
159 - Fixed checking for valid OSes in instance creation
160 - Fixed handling of the "vcpus" parameter in instance listing and in
161 general of invalid parameters
162 - Fixed http server library, and thus RAPI, to handle invalid
163 username/password combinations correctly; this means that now they
164 report unauthorized for queries too, not only for modifications,
165 allowing earlier detect of configuration problems
166 - Added a new "role" node list field, equivalent to the master/master
167 candidate/drained/offline flags combinations
168 - Fixed cluster modify and changes of candidate pool size
169 - Fixed cluster verify error messages for wrong files on regular nodes
170 - Fixed a couple of issues with node demotion from master candidate role
171 - Fixed node readd issues
172 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
173 - Added a new ``--no-voting`` option for masterfailover to fix failover
174 on two-nodes clusters when the former master node is unreachable
175 - Added instance reinstall over RAPI
181 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
182 allow re-adding the start in single-user option (regression from 1.2)
183 - the watcher writes the instance status to a file, to allow monitoring
184 to report the instance status (from the master) based on cached
185 results of the watcher's queries; while this can get stale if the
186 watcher is being locked due to other work on the cluster, this is
188 - the watcher now also restarts the node daemon and the rapi daemon if
190 - fixed the watcher to handle full and drained queue cases
191 - hooks export more instance data in the environment, which helps if
192 hook scripts need to take action based on the instance's properties
193 (no longer need to query back into ganeti)
194 - instance failovers when the instance is stopped do not check for free
195 RAM, so that failing over a stopped instance is possible in low memory
197 - rapi uses queries for tags instead of jobs (for less job traffic), and
198 for cluster tags it won't talk to masterd at all but read them from
200 - a couple of error handling fixes in RAPI
201 - drbd handling: improved the error handling of inconsistent disks after
202 resync to reduce the frequency of "there are some degraded disks for
203 this instance" messages
204 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
205 error handling path called a wrong function name)
211 - no changes from rc5
214 Version 2.0 release candidate 5
215 -------------------------------
217 - fix a couple of bugs (validation, argument checks)
218 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
219 - some small improvements to RAPI and IAllocator
220 - make watcher automatically start the master daemon if down
223 Version 2.0 release candidate 4
224 -------------------------------
226 - change the OS list to not require locks; this helps with big clusters
227 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
228 volume group is broken
229 - ``gnt-instance info``, without any arguments, doesn't run for all
230 instances anymore; either pass ``--all`` or pass the desired
231 instances; this helps against mistakes on big clusters where listing
232 the information for all instances takes a long time
233 - miscellaneous doc and man pages fixes
236 Version 2.0 release candidate 3
237 -------------------------------
239 - Change the internal locking model of some ``gnt-node`` commands, in
240 order to reduce contention (and blocking of master daemon) when
241 batching many creation/reinstall jobs
242 - Fixes to Xen soft reboot
243 - No longer build documentation at build time, instead distribute it in
244 the archive, in order to reduce the need for the whole docbook/rst
248 Version 2.0 release candidate 2
249 -------------------------------
251 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
252 - Fix watcher startup sequence, improves the behaviour of busy clusters
253 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
254 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
255 bind address and other places
256 - Some documentation fixes and updates
259 Version 2.0 release candidate 1
260 -------------------------------
262 - More documentation updates, now all docs should be more-or-less
264 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
266 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
268 - Fix an issue related to $libdir/run/ganeti and cluster creation
274 - Xen PVM and KVM have switched the default value for the instance root
275 disk to the first partition on the first drive, instead of the whole
276 drive; this means that the OS installation scripts must be changed
278 - Man pages have been updated
279 - RAPI has been switched by default to HTTPS, and the exported functions
280 should all work correctly
281 - RAPI v1 has been removed
282 - Many improvements to the KVM hypervisor
283 - Block device errors are now better reported
284 - Many other bugfixes and small improvements
290 - Version 2 is a general rewrite of the code and therefore the
291 differences are too many to list, see the design document for 2.0 in
292 the ``doc/`` subdirectory for more details
293 - In this beta version there is not yet a migration path from 1.2 (there
294 will be one in the final 2.0 release)
295 - A few significant changes are:
297 - all commands are executed by a daemon (``ganeti-masterd``) and the
298 various ``gnt-*`` commands are just front-ends to it
299 - all the commands are entered into, and executed from a job queue,
300 see the ``gnt-job(8)`` manpage
301 - the RAPI daemon supports read-write operations, secured by basic
302 HTTP authentication on top of HTTPS
303 - DRBD version 0.7 support has been removed, DRBD 8 is the only
304 supported version (when migrating from Ganeti 1.2 to 2.0, you need
305 to migrate to DRBD 8 first while still running Ganeti 1.2)
306 - DRBD devices are using statically allocated minor numbers, which
307 will be assigned to existing instances during the migration process
308 - there is support for both Xen PVM and Xen HVM instances running on
310 - KVM virtualization is supported too
311 - file-based storage has been implemented, which means that it is
312 possible to run the cluster without LVM and DRBD storage, for
313 example using a shared filesystem exported from shared storage (and
314 still have live migration)
320 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
321 - Reuse the old instance mac address by default on instance import, if
322 the instance name is the same.
323 - Handle situations in which the node info rpc returns incomplete
325 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
326 - Improved version of batcher:
329 - instance mac address support
330 - support for HVM clusters/instances
332 - Add an option to show the number of cpu sockets and nodes in
334 - Support OSes that handle more than one version of the OS api (but do
335 not change the current API in any other way)
336 - Fix ``gnt-node migrate``
337 - ``gnt-debug`` man page
338 - Fixes various more typos and small issues
339 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
345 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
346 type of disk exported to fully virtualized instances.
347 - provide access to the serial console of HVM instances
348 - instance auto_balance flag, set by default. If turned off it will
349 avoid warnings on cluster verify if there is not enough memory to fail
350 over an instance. in the future it will prevent automatically failing
351 it over when we will support that.
352 - batcher tool for instance creation, see ``tools/README.batcher``
353 - ``gnt-instance reinstall --select-os`` to interactively select a new
354 operating system when reinstalling an instance.
355 - when changing the memory amount on instance modify a check has been
356 added that the instance will be able to start. also warnings are
357 emitted if the instance will not be able to fail over, if auto_balance
359 - documentation fixes
360 - sync fields between ``gnt-instance list/modify/add/import``
361 - fix a race condition in drbd when the sync speed was set after giving
362 the device a remote peer.
368 - note: the allowed size and number of tags per object were reduced
369 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
370 - fixed twisted 8.x compatibility
371 - fixed ``gnt-instance replace-disks`` with iallocator
372 - add TCP keepalives on twisted connections to detect restarted nodes
373 - disk increase support, see ``gnt-instance grow-disk``
374 - implement bulk node/instance query for RAPI
375 - add tags in node/instance listing (optional)
376 - experimental migration (and live migration) support, read the man page
377 for ``gnt-instance migrate``
378 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
379 has some small improvements in handling its state file
385 - Experimental readonly, REST-based remote API implementation;
386 automatically started on master node, TCP port 5080, if enabled by
387 ``--enable-rapi`` parameter to configure script.
388 - Instance allocator support. Add and import instance accept a
389 ``--iallocator`` parameter, and call that instance allocator to decide
390 which node to use for the instance. The iallocator document describes
391 what's expected from an allocator script.
392 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
393 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
394 node is lost there is still enough memory to fail over the instances
396 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
397 ``gnt-cluster verify``, to check for site-specific compliance. All the
398 hooks will run, and their output, if any, will be displayed. Any
399 failing hook will make the verification return an error value.
400 - ``gnt-cluster verify`` now checks that its peers are reachable on the
401 primary and secondary interfaces
402 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
403 that is still declared as part of the cluster and has failed.
404 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
405 specifying output fields, that just adds the chosen fields to the
407 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
408 export from the filesystem.
409 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
410 have been added. Using them you can enable/disable acpi and pae
411 support, and specify a path for a cd image to be exported to the
412 instance. These parameters as the name suggest only work on HVM
414 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
415 PAE support will be set to the previously hardcoded values, but the
416 (previously hardcoded) path to the CDROM ISO image will be unset and
417 if required, needs to be set manually with ``gnt-instance modify``
419 - The address to which an instance's VNC console is bound is now
420 selectable per-instance, rather than being cluster wide. Of course
421 this only applies to instances controlled via VNC, so currently just
422 applies to HVM clusters.
428 - more tweaks to the disk activation code (especially helpful for DRBD)
429 - change the default ``gnt-instance list`` output format, now there is
430 one combined status field (see the manpage for the exact values this
432 - some more fixes for the mac export to hooks change
433 - make Ganeti not break with DRBD 8.2.x (which changed the version
434 format in ``/proc/drbd``) (issue 24)
435 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
436 template, allowing migration from DRBD0.7+MD to DRBD8
442 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
444 - add command aliases infrastructure and a few aliases
445 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
446 man pages and the ``--help`` option of ``gnt-node
447 list``/``gnt-instance list``
448 - fix ``gnt-backup list`` with down nodes (issue 21)
449 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
450 - fix the dist archive and add a check for including svn/git files in
452 - some developer-related changes: improve the burnin and the QA suite,
453 add an upload script for testing during development
459 - experimental HVM support, read the install document, section
460 "Initializing the cluster"
461 - allow for the PVM hypervisor per-instance kernel and initrd paths
462 - add a new command ``gnt-cluster verify-disks`` which uses a new
463 algorithm to improve the reconnection of the DRBD pairs if the device
464 on the secondary node has gone away
465 - make logical volume code auto-activate LVs at disk activation time
466 - slightly improve the speed of activating disks
467 - allow specification of the MAC address at instance creation time, and
468 changing it later via ``gnt-instance modify``
469 - fix handling of external commands that generate lots of output on
471 - update documentation with regard to minimum version of DRBD8 supported
477 - Log the ``xm create`` output to the node daemon log on failure (to
478 help diagnosing the error)
479 - In debug mode, log all external commands output if failed to the logs
480 - Change parsing of lvm commands to ignore stderr
486 - Another round of updates to the DRBD 8 code to deal with more failures
487 in the replace secondary node operation
488 - Some more logging of failures in disk operations (lvm, drbd)
489 - A few documentation updates
496 - Change configuration file format from Python's Pickle to JSON.
497 Upgrading is possible using the cfgupgrade utility.
498 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
499 faster replace disks and is more stable (DRBD 8 has many improvements
500 compared to DRBD 0.7)
501 - Added command line tags support (see man pages for ``gnt-instance``,
502 ``gnt-node``, ``gnt-cluster``)
503 - Added instance rename support
504 - Added multi-instance startup/shutdown
505 - Added cluster rename support
506 - Added ``gnt-node evacuate`` to simplify some node operations
507 - Added instance reboot operation that can speedup reboot as compared to
509 - Soften the requirement that hostnames are in FQDN format
510 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
512 - Removed dependency on debian's patched fping that uses the
513 non-standard ``-S`` option
514 - Now the OS definitions are searched for in multiple, configurable
515 paths (easier for distros to package)
516 - Some changes to the hooks infrastructure (especially the new
517 post-configuration update hook)
518 - Other small bugfixes
520 .. vim: set textwidth=72 :