Minimum and maximum instance specs are put together into a single element
of the instance policy. This is in preparation for introducing multiple
min/max specs.

Signed-off-by: Bernardo Dal Seno <bdalseno@google.com>
Reviewed-by: Helga Velroyen <helgav@google.com>

If there are errors in the configuration when master daemon is started,
they are now logged and reported. Previously they were logged sometimes at
startup, and when performing any change to the configuration.

Signed-off-by: Bernardo Dal Seno <bdalseno@google.com>
Reviewed-by: Helga Velroyen <helgav@google.com>

The code relative to instance specs has been moved into a different
function. This in preparation for further patches to have a more clear code
structure.

Signed-off-by: Bernardo Dal Seno <bdalseno@google.com>
Reviewed-by: Helga Velroyen <helgav@google.com>

Instance policy fields were manually filled in CreateIPolicyFromOpts() and
then filled again by the caller. Now they are filled once inside the
function.

Signed-off-by: Bernardo Dal Seno <bdalseno@google.com>
Reviewed-by: Helga Velroyen <helgav@google.com>

The code is more modular and the output is YAML-compliant. QA updated to
take advantage of it.

Signed-off-by: Bernardo Dal Seno <bdalseno@google.com>
Reviewed-by: Helga Velroyen <helgav@google.com>

The code is more modular and the output is YAML-compliant.

Added function in QA that uses PyYAML to parse the command output, and QA
is updated to take advantage of it (instead of using lots of complicated
REs).

Signed-off-by: Bernardo Dal Seno <bdalseno@google.com>
Reviewed-by: Helga Velroyen <helgav@google.com>

The function produces a human-readable output, which is also a valid YAML
file, from an intermediate data structure. This will be used by "gnt-xxx
info" commands. The output of "gnt-xxx info" was almost YAML-compliant, and
only minor adjustments are needed to make it fully compliant.

Signed-off-by: Bernardo Dal Seno <bdalseno@google.com>
Reviewed-by: Helga Velroyen <helgav@google.com>

The "queryargs" parameter of the __init__ function of rlib2 classes should
be a dictionary, as defined in the ResourceBase parent class (in
lib/rapi/baserlib.py).

In the rlib2 and baserlib unit tests, when the queryargs parameter is not
actually needed, it is sometimes erroneously initialized with an empty list or a
"None" value instead of an empty dictionary.

This commit solves the problem and introduces an assertion to prevent it from
happening again in the future.
The use of the assertion is safe, because in production code the __init__
function is only called by the RAPI server, with the queryargs parameter
initialized by the Mapper.getController function (lib/rapi/connector.py) that
always returns a dictionary. It can only affect test code and future code.

Signed-off-by: Michele Tartara <mtartara@google.com>
Reviewed-by: Guido Trotter <ultrotter@google.com>

cgi.parse_qs() is deprecated in python 2.6 and substituted by
urlparse.parse_qs().

This patch modifies the code to use the new function.

Signed-off-by: Michele Tartara <mtartara@google.com>
Reviewed-by: Helga Velroyen <helgav@google.com>

This fixes the currently broken Q&A. The consistency check
for storage types is now only run when the 'enabled_storage_types'
are actually changed by 'gnt-cluster modify'.

Signed-off-by: Helga Velroyen <helgav@google.com>
Reviewed-by: Guido Trotter <ultrotter@google.com>

When disk templates get disabled via 'gnt-cluster modify', so far
we did not check if any instances still use a disk template of that
type. This patch adds the necessariy check.
It also includes adding a map of disk templates to storage types
and reordering of storage types and disk templates in alphabetical order.

Signed-off-by: Helga Velroyen <helgav@google.com>
Reviewed-by: Guido Trotter <ultrotter@google.com>

This implements an online update of the configuration for when a configuration
is loaded that does not yet contain the 'enabled_storage_types' attribute.
Note that this will require more changes as dis/enabling of (shared) file
storage at configure time will be removed.
This patch includes adding a mapping from disk templates to storage types. While
I was on it, I also reordered the disk template and storage type parameters
alphabetically, to make dealing with those easier.
Also, this mentions the configuration change in the NEWS file.

Signed-off-by: Helga Velroyen <helgav@google.com>
Reviewed-by: Guido Trotter <ultrotter@google.com>

This patch extends the 'gnt-cluster init' command to set the list
of enabled storage types on cluster creation.

Signed-off-by: Helga Velroyen <helgav@google.com>
Reviewed-by: Guido Trotter <ultrotter@google.com>

This patch extends the 'gnt-cluster modify' command to manipulate the list
of enabled storage types. Note that this currenlty does no validation
with respect to whether or not there are instances currently using a storage
type that is being removed from the list.

Signed-off-by: Helga Velroyen <helgav@google.com>
Reviewed-by: Guido Trotter <ultrotter@google.com>

This extends the 'gnt-cluster info' command to list the storage types
that are enabled on the cluster. It also fixes the broken indentation
in the 'handleCall' function.

Signed-off-by: Helga Velroyen <helgav@google.com>
Reviewed-by: Guido Trotter <ultrotter@google.com>

This patch adds the cluster's new field 'enabled_storage_types'
to the configuration objects in python and haskell.

Signed-off-by: Helga Velroyen <helgav@google.com>
Reviewed-by: Guido Trotter <ultrotter@google.com>

This patch adds the currently known and used storage methods to the
constants.py file. Note that storage types are different from disk
templates. For example both disk templates, drbd and plain, use lvm
storage types.

The rapi documentation asserts that valid storage types are limited
to lvm and file. With the extension of storage types, this is not true
anymore, but the implementation of listing storage units is not
implemented for the new types. We disable the assert for now, adding
a fixme to extend it once the listing of storage units is extended
for the new types as well.

'gnt-node list-storage' currently does not support any other storage
type than 'lvm' and 'file'. We limit the QA to storage types that are
currently implemented for this operation.

Signed-off-by: Helga Velroyen <helgav@google.com>
Reviewed-by: Guido Trotter <ultrotter@google.com>

This commit adds the core infrastructure of the monitoring daemon,
and integrates it in the build and test systems.

The actual functionality of the monitoring daemon is still completely
missing.

Signed-off-by: Michele Tartara <mtartara@google.com>
Reviewed-by: Iustin Pop <iustin@google.com>

We are not ready for this change yet. Let's push it to 2.8.

Signed-off-by: Michele Tartara <mtartara@google.com>
Reviewed-by: Guido Trotter <ultrotter@google.com>

Signed-off-by: Helga Velroyen <helgav@google.com>
Reviewed-by: Iustin Pop <iustin@google.com>

Also, add its logfiles and extra log files.

Signed-off-by: Michele Tartara <mtartara@google.com>
Reviewed-by: Iustin Pop <iustin@google.com>

Some daemons will need more than the single logfile that is currently
allowed. This patch introduces the infrastructure to allow this.

Signed-off-by: Michele Tartara <mtartara@google.com>
Reviewed-by: Iustin Pop <iustin@google.com>

This commit adds the user and group variables for the monitoring agent in all
the configuration files where they are required.

Signed-off-by: Michele Tartara <mtartara@google.com>
Reviewed-by: Guido Trotter <ultrotter@google.com>

This allows migrating to any node, as it is already possible for
failover, when instances are externally mirrored.

Signed-off-by: Daniel Krambrock <danielk_lists@z9d.de>
Signed-off-by: Guido Trotter <ultrotter@google.com>
Reviewed-by: Guido Trotter <ultrotter@google.com>

This addresses Issue 237.

Mirroring no disk is a no-op. As such we can treat them like mirrored
instances, since the data they need (none) will be present on all nodes.

This is definitely enough to failover or migrate instances with a manual
target node. Further work might be needed for support in hail.

Signed-off-by: Guido Trotter <ultrotter@google.com>
Reviewed-by: Iustin Pop <iustin@google.com>

This fixes issue 396.

- Fix a wrong comment that mentions drbd8 when actually the code acts
  only on externally mirrored instances.
- Fix a wrong assert that requires failover/migrate to acquire the NAL
  on externally mirrored instances: this is the case only when a
  iallocator is used, not if a node is manually specified.

Extra qa should be added, but this can only be done if an externally
mirrored instance or a diskless instance is tested. It is easier to add
it on the master branch, where we have diskless instances support.

Signed-off-by: Guido Trotter <ultrotter@google.com>
Reviewed-by: Iustin Pop <iustin@google.com>

Now that all fields are implemented, and (I think) behave equivalent
to the Python implementation, we can switch on the split queries for
networks.

Signed-off-by: Iustin Pop <iustin@google.com>
Reviewed-by: Helga Velroyen <helgav@google.com>

Currently, both the Python and Haskell code return the internal
instance list unsorted, which means the output can vary depending on
the phase of the moon (well, the Haskell code actually uses internally
a tree, sorted by the instance name, but it's implementation detail).

By adding an explicit sort step, we guarantee both stability of the
output and consistency between the two query paths.

Signed-off-by: Iustin Pop <iustin@google.com>
Reviewed-by: Helga Velroyen <helgav@google.com>

And also use it for simple network fields. Currently the Python
library returns jsnull, which is formatted as 'None' by the Python
client, which (IMHO) is not nice, since these fields are strings (some
of them) or "other". An undeclared gateway is easier to diagnose when
properly reported by the server as missing, rather than "existing but
null".

Signed-off-by: Iustin Pop <iustin@google.com>
Reviewed-by: Helga Velroyen <helgav@google.com>

Tests for:
  objects.InstancePolicy.CheckParameterSyntax()
  objects.InstancePolicy.CheckDiskTemplates()
  objects.InstancePolicy.CheckISpecSyntax()

Instance policies with an empty disk-template list now are reported.

Signed-off-by: Bernardo Dal Seno <bdalseno@google.com>
Reviewed-by: Guido Trotter <ultrotter@google.com>

IPOLICY_DEFAULTS is now a legal policy (the disk-templates entry was a set
instead of a list, before).

Signed-off-by: Bernardo Dal Seno <bdalseno@google.com>
Reviewed-by: Guido Trotter <ultrotter@google.com>

Unknown elements were silently removed on startup. This means that a
software upgrade could result in lost configuration information if
cfgupgrade wasn't run promptly.

Added unit test for Cluster.UpgradeConfig() to cover this case.

Signed-off-by: Bernardo Dal Seno <bdalseno@google.com>
Reviewed-by: Guido Trotter <ultrotter@google.com>

Policy violations of back-end parameters that used the cluster default
value were not reported in cluster-verify.

Signed-off-by: Bernardo Dal Seno <bdalseno@google.com>
Reviewed-by: Guido Trotter <ultrotter@google.com>

"default" was not accepted as a valid input value for instance specs in
group policies, due to a bug introduced in 2cc673a3 (and released with
2.6.0). Added QA for this and another similar case.

Signed-off-by: Bernardo Dal Seno <bdalseno@google.com>
Reviewed-by: Guido Trotter <ultrotter@google.com>

Instance disk template is checked against the policy, and diskless
instances aren't checked for the number of disks.

Signed-off-by: Bernardo Dal Seno <bdalseno@google.com>
Reviewed-by: Guido Trotter <ultrotter@google.com>

A fragment in LUInstanceCreate.CheckPrereq() removed in commit ba147ff8 was
reintroduced in merge 8e09e801 due to a change in df28c49b.

Signed-off-by: Bernardo Dal Seno <bdalseno@google.com>
Reviewed-by: Guido Trotter <ultrotter@google.com>

There was a typo which prevented the correct option from being passed to
RAPI

Signed-off-by: Daniel Krambrock <danielk_lists@z9d.de>
Signed-off-by: Guido Trotter <ultrotter@google.com>
Reviewed-by: Guido Trotter <ultrotter@google.com>

The "reason" opcode parameter can now apply to multiple opcodes, not just the
one for rebooting instances.

Signed-off-by: Michele Tartara <mtartara@google.com>
Reviewed-by: Iustin Pop <iustin@google.com>

When an OpCode's parameter has a tuple as default value, this code
will break:

  buf.write("defaults to ``%s``" % default)

The patch fixes this and other potential cases by always passing a
tuple to '%'.

Signed-off-by: Iustin Pop <iustin@google.com>
Reviewed-by: Michele Tartara <mtartara@google.com>

The "reason" attribute will be common to many opcodes.
This patch factors it out so that it will be possible for the next commits
to use it without code duplication.

Signed-off-by: Michele Tartara <mtartara@google.com>
Reviewed-by: Guido Trotter <ultrotter@google.com>