Unit tests for spindles and exclusive storage in htools
Everything should work the same when exclusive storage is active, asspindles are not yet handled as a resource. A new test has been added forthe interactions between spindles and policies.
Signed-off-by: Bernardo Dal Seno <bdalseno@google.com>...
Check real spindles in ipolicies
When exclusive storage is enabled, the spindles in instance disks are usedto check the instance policies (as outlined in design-partitioned.rst).
Signed-off-by: Bernardo Dal Seno <bdalseno@google.com>Reviewed-by: Klaus Aehlig <aehlig@google.com>
Add spindles to instance disks in htools
A new data type is introduced for disks to store both size and spindles.When available, spindles are filled with input data. Except for loading andstoring, spindles are ignored.
Load exclusive_storage in htools
The node parameter is loaded into the data structures. No behavior is yetmodified.
New function to load JSON arrays of optional values
This will be needed to load spindles in some htools backends. Unit testsprovided.
Add unit test for text backend + fix bug
Test serialization and deserialization of instances. Fix check of secondarynode.
Test with non-redundant instances and several node-groups
The example cluster contains two node groups. The first contains three nodes,with two non-redundant instances each. As each node has only capacity for threeinstances, the nodes have to be rebooted individually. Additionally, there is...
Merge branch 'stable-2.8' into master
Add tests for network-aware allocation
hail-alloc-invalid-network defines a cluster with two nodegroups and anallocation request which does not fit on any of the groups. Group 1 hasinvalid disk-templates while Group 2 is not connected to the rightnetworks....
Parse NIC data from allocation request in hail
Add a NIC type and extend the Instance type by a list of NIC's. Parsethe NIC's in allocation requests and store them for now. Later patcheswill make use of this field in order to ensure that the requestedinstance is only placed in node groups wich are connected to those...
Support group networks in Text backend
The Text backend now parses network UUID (comma separated) andserializes them in the same form.The test data is adapted to the new format.
Signed-off-by: Thomas Thrainer <thomasth@google.com>Reviewed-by: Guido Trotter <ultrotter@google.com>
Parse node group networks
Extend the Group by the network ids it is connected to. Adaptthe IAlloc backend such that the networks are parsed correctly.This also required the adaption of test data.
Make the disks_active flag queryable
gnt-instance list now also supports the disks_active field.
Signed-off-by: Thomas Thrainer <thomasth@google.com>Reviewed-by: Helga Velroyen <helgav@google.com>
Add disks_active to configuration
This flag tracks if the disks of an instace are supposed to be active.That's the case when an instance is running or when its disks gotactivated explicitly (and in a couple of other cases).It will be used by watcher to re-activate disks after a node reboot....
Attaching a logical volume builds a list of PVs
When an LV gets attached, the list of the PVs used by the LV is built. Thiswill be used to count spindles for exclusive_storage, but it could also beuseful to optimize disk growing and snapshotting.
Refactor code for attaching a logical volume
The parsing of "lvs" output is moved into private methods. The code isslightly more readable and testable. The split in two methods is usefulfor the following patches. Unit tests for the new functions areprovided....
Merge branch 'stable-2.8'
Add spindles field to disk object
The field is filled with the value provided on the command line.
Signed-off-by: Bernardo Dal Seno <bdalseno@google.com>Reviewed-by: Thomas Thrainer <thomasth@google.com>
cfgupgrade: Downgrade is a NO-OP
The configuration is still the same as in 2.8 (the reference stable versionfor this branch), so downgrade shouldn't do anything.
Unit tests are also updated, with a new 2.8 configuration file. Theconfiguration file used for the upgrade+downgrade test was tailored to the...
Extend hroller tests by options for non-redundant instances
The cluster now consists of 3 nodes, with drbd instances betweennodes 1 and 2, and 2 and 3. Additionally, nodes 1 and 3 each containa non-redundant instance, but node 2 cannot hold two additional...
Test for hroller taking non-redundant instances into account
The example cluster consists of 6 nodes, each hosting 2 instances andhaving capacity for 3. So, while the drbd-induced graph consists ofonly insulated nodes, no more than two nodes can be rebooted at the...
Stash Xen config file after a failed startup
This is a fix for a minor bug.Currently, a failed Xen VM start results in a stale config file leftbehind on the filesystem.This change introduces a new log directory, where the Xen VM configfile is moved after a failed startup....
cmdlib: Cleanup public/private functions
All functions/classes which are used outside of their defining module(with tests as an exception) no longer have a leading underscore.
Signed-off-by: Thomas Thrainer <thomasth@google.com>Reviewed-by: Bernardo Dal Seno <bdalseno@google.com>
cmdlib: Extract storage related functionality
Split instance.py further by extracting storage related logical unitsand functions to instance_storage.py.
Extract query related logical units from cmdlib
All LUQuery* classes are extracted to query.py.
Extract instance related logical units from cmdlib
All LUInstance* classes are extracted to instance.py. Common functionsare moved to common.py if used by non-instance logical units as well.Additionally, helper functions which are only used by LUBackup* and...
Extract node related logical units from cmdlib
All LUNode* classes are extracted to node.py. Common functions are movedto common.py if used by non-node logical units as well.
Extract group related logial units from cmdlib
All LUGroup* classes are moved to group.py. Common functions areextracted to common.py.
Extract cluster related logical units from cmdlib
All LUCluster* classes are extracted to cluster.py. Shared functions areextracted to common.py, helper functions only used by LUCluster* areextracted to cluster.py.
Signed-off-by: Thomas Thrainer <thomasth@google.com>...
Compatibility test for instances
This patch introduces a test to check the compatibilityof the Haskell and the Python representation of instances.
Signed-off-by: Helga Velroyen <helgav@google.com>Reviewed-by: Klaus Aehlig <aehlig@google.com>
Instance generators
This patch introduces and enhances generators forinstances:- 'genInstWithNets' is split into the generation of an arbitrary instance and enhancing an instance with nets- 'genInst' calls 'genInstWithNets' with an empty set of initial networks to provide a reasonable default...
Annotate every arbitrary instance field
The Arbitrary instance of the 'Instance' object is writtenusing the <*> syntax. Since it often uses the 'arbitrary'generator for the instance's fields it is hard to figureout which 'arbitrary' fills which instance field. This...
Generators for disks
This patch adds generators for Disk instances to the Haskelltest code. It uses somewhat more reasonable generators tofill the fields instead of just arbitrary values.'genDiskWithChildren' is a generator that generates a diskwith a specified number of disk children. To avoid shooting...
Use os.statvfs to determine free disk space
This simplifies my previous commit (820bade90) by using os.statvfsinstead of parsing the output of 'df'.
Signed-off-by: Helga Velroyen <helgav@google.com>Reviewed-by: Bernardo Dal Seno <bdalseno@google.com>
Backend function for file storage space reporting
This adds functionality to retrieve disk space informationfor file storage. It calls the 'df' tool and parses itsoutput to extract the total and free amount of disk spaceon the disk where the given path is located....
Move 'container.py' to storage directory
Moving 'container.py' to the storage directory.
Signed-off-by: Helga Velroyen <helgav@google.com>Reviewed-by: Thomas Thrainer <thomasth@google.com>
Rename dir 'block' to 'storage'
Renaming the 'block' directory to 'storage', because I plan toplace code there that is related to file storage and leavingit named 'block' would be misleading.
Rename storage.py to container.py
Renaming 'storage.py' to 'container.py'. It will be moved into the new'storage' directory, which will come in later patches to avoid clashes ofnotation.
Extend hroller test to also verify tag-based node selection
While the multiple-tags test was added to verify that coloring is doneonly after node selection (otherwise it wouldn't be possible to get inboth cases a single reboot group), it can easily be extended to also...
Add a test for online rolling reboot scheduling
In the example configuration, the graph constructed by just connectingprimary and secondary instances is two-colorable. However, when takingconflicting locations of secondary nodes into account, three reboot...
Fix expectation in hroller test
Regular expressions are not shell globs. So "any symbol" is expressedby a dot, not a question mark. In this case, the confusion lead to atoo liberal expectation, hence the test passed. Fix it nevertheless.
Signed-off-by: Klaus Aehlig <aehlig@google.com>...
Add a test demonstrating the --node-tags option of hroller
The example is a cluster of 6 nodes, paired into 3 group by threeinstances. So the whole cluster would need two reboot groups. The twotags select, in two different ways, one node of each group. So, when...
Add tests for the -O option of hroller
In hroller, the option -O can be used to mark certain nodes as offline.These nodes should then not be part of any reboot group. Add teststo verify this behavior.
Signed-off-by: Klaus Aehlig <aehlig@google.com>Reviewed-by: Guido Trotter <ultrotter@google.com>
Split DRBD8Dev into DRBD8 and DRBD8Dev
All functionality specific to a single DRBD8 devide is now in DRBD8Dev,whereas functionality which is valid for the whole DRBD "installation" on a device is collected in DRBD8.
This makes it possible to remove a couple of FIXME's and clarifies the...
Rename DRBD8 to DRBD8Dev
Right now the DRBD8 class has multiple responsibilities: a) itreprensents a device which can be set up, grown, etc. and b) itrepresents the whole DRBD system on a node which has a usermode helper,which knows how to shut down all devices, etc. Therefore, the DRBD8Dev...
Make DRBD version queryable from noded
gnt-cluster verify should issue a warning if there are multiple DRBDversions present in a node group. In order to do so, the DRBD versionhas to be queryable from noded.
Merge remote-tracking branch 'origin/stable-2.8'
Conflicts: lib/bdev.py lib/cmdlib.py
Signed-off-by: Michele Tartara <mtartara@google.com>Reviewed-by: Klaus Aehlig <aehlig@google.com>
Disk templates are "allowed" and not "enabled" in ipolicies
The way disk templates were displayed in instance policies became confusingsince the introduction of the enabled_disk_templates cluster parameter.
Fix issue 440.
Unit test for FillIPolicy keeping unknown keys
This makes sure that issue 401 doesn't come back.
Signed-off-by: Bernardo Dal Seno <bdalseno@google.com>Reviewed-by: Helga Velroyen <helgav@google.com>
Add reason trail to haskell opcode format
The haskell type definition of opcodes should remain aligned with the pythonone.
Signed-off-by: Michele Tartara <mtartara@google.com>Reviewed-by: Helga Velroyen <helgav@google.com>
Add command generator for DRBD 8.4
The DRBD84CmdGenerator class, which generates commands suited for DRBD8.4, is added. A common baseclass for DRBD83CmdGenerator andDRBD84CmdGenerator is introduced as well.
Add test for version dispatching code
Based on the version of DRBD (as reported through /proc/drbd), the DRBD8class chooses helper objects which adapt for the differences betweenDRBD <= 8.3 and DRBD 8.4. This new test verifies that the proper objectsare created....
Extract DRBD related tests
As the DRBD code was extracted from bdev.py, the corresponding testsshould as well be extracted into their own file. That's what this patchperforms.
Signed-off-by: Thomas Thrainer <thomasth@google.com>Signed-off-by: Michele Tartara <mtartara@google.com>...
Extract command generation logic from DRBD8
In order to easily support the DRBD 8.4 command syntax, this patchextracts all command generation logic (i.e. where calls to `drbdsetup`are assembled) into a new class DRBD83CmdGenerator.
Add `drbdsetup show` parser for DRBD 8.4
Common functionality between the DRBD 8.3 and DRBD 8.4 parser has beenextracted into BaseShowInfo. A test which verifies the behaviour isincluded, but the DRBD84ShowInfo class is not yet used within the DRBD8class....
Extract DRBD info parsing into drbd_info.py
Parsing the status (/proc/drbd) and `drbdsetup show` output takes quitea bit of code, which has to be augmented by even more with the DRBD 8.4support. So extract all the related classes into their own file.
Extract DRBD8ShowInfo class
This class parses the `drbdsetup show` output and represents it in aneasily accessible format. It got extracted so that 1) the DRBD8 classcan focus more on DRBD logic rather than parsing and 2) it's easier toadapt the parser to new formats....
Add test for DRBD 8.4 /proc/drbd parsing
The current parser already supports the output of DRBD 8.4 /proc/drbd,so add tests which actually verify this behaviour.Note that the empty lines in proc_drbd84.txt are actually present in theoutput of DRBD 8.4, they always appear when minors are not contiguous....
Add test for empty version in /proc/drbd
The previous patch removed a (rather implicit) test for /proc/drbd notcontaining a version line. This adds an explicit test for this case.
Extract /proc/drbd parsing code into DRBD8Info
As the DRBD8 class got bigger due to the previous merge of BaseDRBD, nowparts of it are ripped out into DRBD8Info. This new class parses/proc/drbd and exposes the information in an easily accessible way. This...
Remove BaseDRBD
BaseDRBD was probably useful when DRBD 0.7 and DRBD 8 were supported.However, there is only one subclass of BaseDRBD remaining (DRBD8), andthe separation of responsibilities between those two classes was ratherrandomly chosen.
The unification into one class also eases the introduction of...
Rename _MassageProcData to _JoinProcDataPerMinor
That's what the method actually does, so there is no reason why weshould use a funny but unreadable name.
Add shelltests for hspace allocation
Both tiered and standard allocations are tested, with a single and a doublemin/max instance specification.
Unit test for cli.FormatPolicyInfo()
We try to parse the string obtained by formatting a policy, and see if weget the same policy.
Add command-line support for multiple specs in ipolicy
Command line options accept multiple min/max specifications in instancepolicies. The output of show-ispecs-cmd is updated accordingly.
Add multiple min/max specs in instance policy
Now instance policies can contain more than one min/max specs. This is themain element of the "Constrained instance sizes" section in the"Partitioned Ganeti" design doc.
This is a big patch, but changing the type of a configuration item requires...
Separate checks for std spec compliance
This is needed to be able to validate the std spec against multiple min/maxspec pairs (appearing in next patches).
Merge branch 'stable-2.7' into stable-2.8
Conflicts: NEWS doc/rapi.rst
Reason trail implementation for "start"
Opcode-specific implementation of the reason trail for the instancestartup operation.
Signed-off-by: Michele Tartara <mtartara@google.com>Reviewed-by: Guido Trotter <ultrotter@google.com>
Reason trail implementation for "shutdown"
Opcode-specific implementation of the reason trail for the instance shutdownoperation.
Add unit tests for cfgupgrade with a real configuration
Test upgrade and downgrade from a realistic 2.7 configuration.
Split functions in cfupgrade unit tests
This makes easier to test real configurations.
Limit specs in ipolicy command lines are always complete
Command line options are brought in line with the specs change of previouspatch. Old options are still allowed in gnt-cluster init, where thesemantic will remain non-ambiguous even after introducing multiple specs....
Refactor tests in ipolicy validation
This makes the code more modular, in preparation for the following patches.
Limit specs in instance policies are always complete
Specs used to specify limits in instance policies are always complete, evenat group level (only the whole limit set can be overridden). This is inpreparation for introducing multiple limits.
Add --ipolicy-xxx-specs options
These options allow to specify whole instance policy specs. This is neededfor the upcoming changes that tend to threat specs as monolithic objects.
Add command to print ipolicy options
The output of this command can be used to create an exact copy of thecurrent instance policy specs. The command could be expanded to print allthe options used to create a group or the cluster.
New CLI input type: list of key/value pairs
This will be used for the new instance specs options.
Allow "default" as a value for ipolicy templates
Previously the string value was being corrupted.
Extract DRBD-related classes into block/drbd.py
This is in preparation to introducing support for DRBD8.4.base.py had to be extracted as well in order to avoid cylic importsbetween bdev.py and drbd.py. It now contains the BlockDev class andutility functions needed by bdev.py and drbd.py....
Move lib/bdev.py to lib/block/bdev.py
That's in preparation of extracting DRBD related code from bdev.py. Asbdev.py is already rather long, new features will require to split itinto more manageable pieces. That's why it's moved in an own directory.
Extend the simulation backend to also simulate a master node
In a simulated cluster as created by the simulation backend tothe htools, make the first node of the first node group the masternode. In this way, hools (like hroller) that require a master node...
Add unit tests for hroller
Add a unit test for hroller in a situation where the orderis completely determined. There are 3 nodes where node 2 has,for each of the other nodes, an instanced shared with it. Sonode 2 has to be a reboot group of its own. There are no more...
Reason trail implementation for instance reboot
This commits allows ganeti to correctly forward the reason trail informationregarding instance reboot.
Remove old "reason" implementation
Remove the useless parts of the old, partial, implementation of the support fortracking the reason of instances state change, before implementing the newreason trail support, as per the design document.
Signed-off-by: Michele Tartara <mtartara@google.com>...
Make hroller insist on finding precisely one master node
As people rely on the master node being the last node of the lastgroup, make hroller fail, if no master node could be found in thecluster. This happens, e.g., if a backend format is used that does not...
Add nic/disk name and UUID fields in Haskell code
Add name and uuid fields to Disk and nic Haskell objects. Also,since they now have an UUID, make them instances of UuidObject.
Signed-off-by: Christor Stavrakakis <cstavr@grnet.gr>Reviewed-by: Helga Velroyen <helgav@google.com>
Add name to INicParams and IDiskParams
This commits adds name to NIC and Disk modification definition.
Signed-off-by: Chris Stavrakakis <cstavr@grnet.gr>Reviewed-by: Helga Velroyen <helgav@google.com>
Update QA tests to check disk and NIC names
Modify existing tests to use the name option for instance disks. Theconfiguration options 'disk' and 'disk-growth' are merged into a single'disks' option, which is a list of dictionaries with 'size', 'growth' and...
Check that device names are unique and valid
Extend the CheckArguments phase of LUInstanceCreate and CheckPrereqphase of LUInstanceSetParams to also check if the name parameters ofdisks and NICs are unique and valid.
Signed-off-by: Christos Stavrakakis <cstavr@grnet.gr>...
Support UUIDs and names when refering to a device
Modify _ApplyContainerMods function to lookup NICs/Disks not only by theirindex inside the container, but also by their UUID or name. Abstract thelookup code in new GetItemFromContainer function.
Make type of identifier in "opcode._TestInstSetParamsModList" to be...
Move HooksMaster out of the mcpu module
We need to do this, so that backend.py doesn't need to import mcpu, andthus indirectly cmdlib. This reduces the size of the node daemon byabout half, which is very important as it is pinned in memory.
This solves Issue 419....
QA: Decision about disk templates to use is more modular
qa_config.IsTemplateSupported() is already used to decide which disktemplates are supported by QA. Enabled templates now are part of thatdecision.
Unit tests are updated.
Adapt test data to changed ipolicy format
The test data for hail allocating a multi-disk instance wasoriginally added in the 2.7 branch. On master the formatof the ipolicy had changed; "min" and "max" are no longertop-level of the ipolicy, but instead below the "minmax"...
Merge branch 'devel-2.7'
Removes obsolete 'enabled storage types'
Since managing of different storage units is now done usingdisk templates and not storage types, we remove the obsoleteenabled storage types.
Signed-off-by: Helga Velroyen <helgav@google.com>Reviewed-by: Michele Tartara <mtarara@google.com>
gnt-cluster modify: modify enabled disk templates
Adds setting of the list of enabled disk templates to 'gnt-cluster modify'.Note that this does not yet include any verification regarding disk templatescurrently used by instances.
Signed-off-by: Helga Velroyen <helgav@google.com>...
Add shelltests verifying hail applies disk ipolicy per disk
We test on two minimal examples, a positive and a negative one, wherethe possibility to allocate a node depends on whether the disk policyis checked on a per-disk level or on the total amount of disk space...