Currently, if converting an instance from plain to drbd fails after
renaming the original LVs, the instance is left in an inconsistent
state. This commit tries to undo the renaming if a failure occurs
on assembling a DRBD device, e.g., when device minor numbers are
exhausted. (Issue 229)

Signed-off-by: Klaus Aehlig <aehlig@google.com>
Reviewed-by: Thomas Thrainer <thomasth@google.com>

Currently this is checked only when distcheck-release is called.
Check it in check-news instead, for all versions before the "current"
one (as defined by configure.ac)

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

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

The renew-crypto entry had a few obscure sentences. Improve it by
rephrasing it.

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

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

- Most entries for beta1 are moved to the top level
- The beta1 to beta2 delta is left

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

During the creation of an instance, if the creation of disks fails, some
partially created disks might remain lying around. There is an already
implemented cleanup procedure, but it requires the disks to be listed in the
disks_created variable, which is updated after the creation of the disk itself
to prevent it from removing disks already existing previously.

This patch introduces a better tracking of partially created disks, so that
they can be removed without risking removing already existing ones.

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

Specify that the need to manually modify the metavg only applies to
clusters with DRBD enabled, and is not always required.

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

The current behaviour of a volume group name change after the cluster
has been initialized can be somewhat unclear. This patch specifies it
more esplicitly in the man page.

Signed-off-by: Michele Tartara <mtartara@google.com>
Reviewed-by: Klaus Aehlig <aehlig@google.com>

The default metavg was always set to be the constant value "xenvg".
This is OK for most cases, but if the cluster is initialized with
a default name through the --vg-name option, the metavg should change
as well.

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

Without this masterd fails saving the config, because it can't change
the group on the file.

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

We need to do this, so that backend.py doesn't need to import mcpu, and
thus indirectly cmdlib. This reduces the size of the node daemon by
about half, which is very important as it is pinned in memory.

This solves Issue 419.

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

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>

Provide some more insights about how to use the RAPI from the command line
through the "curl" command.

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

Add some additional information about how to use/test the RAPI,
including how to specify authentication credentials.

Cherry picked from 681f4f7b on master.

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

Due to a bug in 8e55e20f burnin's installation path got changed.
Moving tools/burnin to the right makefile target reverts this (while
preserving the value of the original commit).

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

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

hail now correctly applies the instance disk policies, which
leads to differnt allocations.

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

We test on two minimal examples, a positive and a negative one, where
the possibility to allocate a node depends on whether the disk policy
is checked on a per-disk level or on the total amount of disk space
requested.

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

In that way, tools building on Instance will benefit from the corrected
verification semantics of the instance policy on disk space.

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

Instance policy on disks is specified on a per-disk basis. So extend
the instance description by the sizes of the individual disks and modify
the instance policy verification to correctly check individual disks.

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

Importing an instance, like JSON, automatically allows the names of all
its functions to be used.

If one of this functions (namely showJSON) is imported separately, GHC 6
is fine with that, but GHC 7 emits a warning (therefore, the compilation of
Ganeti fails because it requires no warnings).

This patch fixes the build for GHC 7 and has no impact on GHC 6.

Note that the same does not apply to the import of readJSON, because that
function is actually used in the same file, so it must be imported separately.

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

When executed in split user mode, Ganeti requires a specific set of users
and groups to be created in the appropriate way.

This commit introduces a tool that takes care of that automatically.

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

The code introduced by the previous commit triggered a possible library
conflict in Ubuntu Lucid.

This patch introduces an equivalent but more widely acceptable version of
the same code.

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

In case of problems while reading the job list from disk (such as permission
errors) confd would silently fail, writing a warning on the log file but
sending an empty list and no error message to the client.

Also, tests have been updated in accordance to the new interface of the modified
functions.

This commit fixes this problem [Issue 405].

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

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

A group name was erroneusly inserted in the list of user names.

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>

Also update the release date for beta2

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

Before:
$ ./autotools/check-news < ./NEWS
Traceback (most recent call last):
  File "./autotools/check-news", line 150, in <module>
    main()
  File "./autotools/check-news", line 120, in main
    parsed_ts = time.mktime(time.strptime(m.group("date"), "%d %b %Y"))
AttributeError: 'NoneType' object has no attribute 'group'

After:
$ ./autotools/check-news < ./NEWS
Line 8: Invalid release line

This happened because the error function just updates an error log,
rather than bailing out immediately.

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