LogicalVolume.Attach in bdev.py is calling "lvs" and specifying a unit of
Megabytes that is then converted to an integer resulting in the value 0 for
small sizes.

This patch makes Ganeti use KB for the unit instead of MB.

Fixes Issue 347.

Signed-off-by: Aaron Knister <aaron.knister@gmail.com>
Reviewed-by: Michele Tartara <mtartara@google.com>

Trasforming the disk of an instance from DRBD to plain did not properly
update the iv_name of disks, leaving it to None.

This commit modifies _ConvertDrbdToPlain to properly set the iv_name
variables.

An analogous fix is not required in the opposite function,
_ConvertPlainToDrbd, because it uses _GenerateDiskTemplate, that already
takes care of doing that.

Fixes Issue 403.

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

This patch introduces an upper limit to the size of the networks that can
be created.

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

When creating a network, so far no size constraints were checked.
We now limit the size of a network to a /30 or bigger, although
tecnically, the ipaddr library supports even /32 networks.

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

If split users are used, the queue directory could only be accessed
by masterd, but also confd needs to be able to read it, e.g. when it
is queried as part of "gnt-job list"

This commit fixes the permissions in such a way to allow proper access rights.

Fixes Issue 406.

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

For diskless instances it is still reasonable to sum up the disk usage
of all the (zero) disks, resulting in the empty sum. This uniformity
also has the advantage that iallocators (like hail) do not have to do
any special considerations for diskless instances.

Signed-off-by: Klaus Aehlig <aehlig@google.com>
Reviewed-by: Helga Velroyen <helgav@google.com>

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

Moving an instance is done by copying over the disks. Restrict
this to disk templates that are copyable. This avoids accidental
use on, e.g., the sharedfile template.

Signed-off-by: Klaus Aehlig <aehlig@google.com>
Signed-off-by: Bernardo Dal Seno <bdalseno@google.com>
Reviewed-by: Bernardo Dal Seno <bdalseno@google.com>

This list contains the disk templates suitable for moving
an instance by copying the files. A requirement is that
they're not accessed externally or shared between nodes;
in particular, sharedfile is not suitable.

Signed-off-by: Klaus Aehlig <aehlig@google.com>
Signed-off-by: Bernardo Dal Seno <bdalseno@google.com>
Reviewed-by: Bernardo Dal Seno <bdalseno@google.com>

Now that _CreateDisks cleans up after itself in case of failure,
do not clean up at call sites, as there we have to overapproximate
thus potentially causing data loss.

Signed-off-by: Klaus Aehlig <aehlig@google.com>
Reviewed-by: Guido Trotter <ultrotter@google.com>

_CreateDisk used to just throw an exception if _CreateBlockDev failed
leaving the caller in the state that some disks were created, without
precise knowledge which. Usually, the clean up then overapproximated
by removing all disks of the instance. This can cause data loss in case
the disk creation failed because of a file being already in the place
where the disk is to be created.

With this patch, upon failure, _CreateDisk removes the disks that were
successfully created, leaving a well-defined state.

Signed-off-by: Klaus Aehlig <aehlig@google.com>
Reviewed-by: Guido Trotter <ultrotter@google.com>

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

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>

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>

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 vcluster changes broke the disabling of file storage; we can
workaround by (manually) skipping the virtualisation of file storage
paths if they are not enabled.

Note that tests/QA are still broken with disabled file storage; this
patch only fixes production code.

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

Currently, the iallocator interface requires LVM storage, due to the
way it computes the node storage information.

By changing the code to understand that GetVGName() can return None,
and by setting the disk_total/disk_free node parameters to the value
zero, we can workaround this issue until proper multi-storage
reporting is implemented.

The hail iallocator plugin works without changes once this is
implemented, as it already handles non-LVM storage as "foreign"
storage (with unlimited capacity).

The patch also changes a bit how the parameters are retrieved and
used; hopefully this is a bit safer and more readable.

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

'MakeLegacyNodeInfo' is not the best place for this, but we'd have to
duplicate it if we wanted a LVM-less version, so the easiest is to add
an optional parameter that allows it to accept/skip LVM-less results.

It still requires at most one VG result, so its behaviour isn't
changed in this respect.

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

This is similar to commit 8775e62a; the addition of node_whitelist
broke this LU as well.

Signed-off-by: Iustin Pop <iustin@google.com>
Reviewed-by: Bernardo Dal Seno <bdalseno@google.com>

Passing --net 0:add,ip=5.5.5.5 failed due to a reference
of a non initialized variable (new_net_obj). Reorder the checks
and add some comments to be readable.

Signed-off-by: Dimitris Aragiorgis <dimara@grnet.gr>
Reviewed-by: Guido Trotter <ultrotter@google.com>

Remove any early returns in LUNetworkDisconnect/LUNetworkConnect
and replace them with if-else statements.

Signed-off-by: Dimitris Aragiorgis <dimara@grnet.gr>
Reviewed-by: Michael Hanselmann <hansmi@google.com>

In this method self.tags might be None and cannot be used
in join(). Use GetTags instead.

Signed-off-by: Dimitris Aragiorgis <dimara@grnet.gr>
Reviewed-by: Guido Trotter <ultrotter@google.com>

networks config objects have UUIDs and thus should be included
in _AllUUIDObjects().

Signed-off-by: Dimitris Aragiorgis <dimara@grnet.gr>
Reviewed-by: Guido Trotter <ultrotter@google.com>

Locks for group instances are aquired only if conflicts are checked.
To this end we must _CheckNodeGroupInstances() only then otherwise
this check will always fail (owned_instances will be []).

Signed-off-by: Dimitris Aragiorgis <dimara@grnet.gr>
Reviewed-by: Guido Trotter <ultrotter@google.com>

Params passed in _CreateNewNic() are not yet evaluated and include
the value passed by user for the network. A lookup must be done first
in order to find the corresponding network UUID which should be stored
in the newly created NIC object.

Params passed in _ApplyNicMods() include the value passed by user
for the network. Thus a lookup must be done first in order to find
the corresponding network UUID. Then compare it with the existing
value and act accordingly.

Signed-off-by: Dimitris Aragiorgis <dimara@grnet.gr>
Reviewed-by: Guido Trotter <ultrotter@google.com>

LUNetworkConnect/LUNetworkDisconnect, in case locking is used, might
lock instances that exist in the requested node group. The acquired
locks should be checked if they are correct at the beginning of
CheckPrereq() via _CheckNodeGroupInstances().

Signed-off-by: Dimitris Aragiorgis <dimara@grnet.gr>
Reviewed-by: Michael Hanselmann <hansmi@google.com>

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

Make LookupNetwork() return None in case target is None. This fixes
Issue 380. Rapi passes network=None and the lookup should not fail.

Make network client aware of new nic.network.gnt-network info shows
the IPs of each instance inside the network.  It parses nic.networks
field of QueryInstances() output, which now is a list of uuids an
not names.

Signed-off-by: Dimitris Aragiorgis <dimara@grnet.gr>
Reviewed-by: Helga Velroyen <helgav@google.com>

In case a NIC is not inside a network then netinfo None. Thus
netinfo["name"] fails.

Signed-off-by: Dimitris Aragiorgis <dimara@grnet.gr>
Reviewed-by: Helga Velroyen <helgav@google.com>

In backend NICs arrive with netinfo filled. If nic.network is not None
nic.netinfo is not too. Thus all the info is derived from HooksDict()
and nic.network must not be checked.

Signed-off-by: Dimitris Aragiorgis <dimara@grnet.gr>
Reviewed-by: Helga Velroyen <helgav@google.com>