For whatever reason, my test cluster managed to acquire
shared_file_storage_dir with a None value, instead of empty
string. This is not flagged in masterd itself, but the node daemon
will fail in writing the value to disk, as it calls len() on the
received value.

Since this is a bad case, we should detect it as soon as possible (we
basically shouldn't be able to set it), but in the meantime we at
least prevent ssconf writes with such values.

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

New versions of KVM support booting from HTTP-hosted ISO
images, via libcurl. This patch adds a proper check to
allow defining either a sane, absolute path or an HTTP
URL as an iso image path.

Remove "format=raw" from the cdrom device options when iso_image
starts with "(https?|ftp)://", to allow latest KVM versions (qemu-kvm) to
boot from HTTP-hosted iso images. Added FILE_OR_URL check to allow
booting from HTTP (kvm).

Signed-off-by: Apollon Oikonomopoulos <apollon@noc.grnet.gr>

Named parameters are only handled by zsh completion for the time being.

Signed-off-by: Apollon Oikonomopoulos <apollon@noc.grnet.gr>

Add completion generator script for zsh. Using native zsh completion functions
we can auto-complete things like instance/node names, job ids, nic/disk/backend
parameters etc.

The generated completion file can be copied to e.g.
/usr/share/zsh/functions/Completion/Unix/_ganeti under Debian, or any other
location under $fpath on other systems.

Signed-off-by: Apollon Oikonomopoulos <apollon@noc.grnet.gr>

Add suggestions for disk-, nic-, and backend-parameter completion.

Also alter autotools/build-bash-completion to ignore the new suggestion types
for the moment.

Signed-off-by: Apollon Oikonomopoulos <apollon@noc.grnet.gr>

Commit faaabe3c fixed failover behaviour for DTS_INT_MIRROR instances, however
it broke migration for DTS_EXT_MIRROR instances, by moving iallocator and node
checks from LUInstanceMigrate to TLMigrateInstance. This has the side-effect
that the LU called the TL with None for both, node and iallocator when the
default iallocator was being used.

This patch maintains the iallocator checks in TLMigrateInstance and fixes the
LU-TL integration.

Signed-off-by: Apollon Oikonomopoulos <apollon@noc.grnet.gr>

Signed-off-by: Apollon Oikonomopoulos <apollon@noc.grnet.gr>

It is now possible to allow adopting a disk during gnt-instance modify time, as
follows:

gnt-instance modify --disk add:adopt=/path/to/disk (blockdev)
  or
gnt-instance modify --disk add:adopt=<lvname> (plain)

We do the same checks as during instance creation.

Signed-off-by: Apollon Oikonomopoulos <apollon@noc.grnet.gr>

Patches db366d9a and aac4511a added support for EXT_MIRROR instances,
but inadvertently introduced a bug: for INT_MIRROR cases, we don't
need (actually we can't support) neither an iallocator nor a target
node.

To fix this, we move the iallocator/node checks in CheckPrereq (or
respectively in the tasklet CheckPrereq), where we have access to the
instance configuration, and additionally we check for and prevent
passing either of these two for INT_MIRROR instances.

Signed-off-by: Iustin Pop <iustin@google.com>
Reviewed-by: René Nussbaumer <rn@google.com>
Reviewed-by: Michael Hanselmann <hansmi@google.com>

DTS_INT_MIRROR better contrasts DTS_EXT_MIRROR.

Signed-off-by: Apollon Oikonomopoulos <apollon@noc.grnet.gr>
[iustin@google.com: updated patch for changed context]
Signed-off-by: Iustin Pop <iustin@google.com>
Reviewed-by: Iustin Pop <iustin@google.com>

Make burnin work with -t sharedfile and update burnin's help.

Shared block storage support is *not* included, as currently there is no way to
provision shared block storage from within Ganeti.

Signed-off-by: Apollon Oikonomopoulos <apollon@noc.grnet.gr>
Signed-off-by: Iustin Pop <iustin@google.com>
Reviewed-by: Iustin Pop <iustin@google.com>

Disable host cache for externally mirrored disks to avoid cache incoherency.
Without this, migrations between the same two nodes may end up in disk
corruption.

This is a runtime override of cluster defaults, mostly a workaround.
Unfortunately, we cannot use cmdlib._CheckHVParams, as we want to check the
hypervisor parameters *in conjunction* with the disk template, which is part of
the instance definition.

Signed-off-by: Apollon Oikonomopoulos <apollon@noc.grnet.gr>
Signed-off-by: Iustin Pop <iustin@google.com>
Reviewed-by: Iustin Pop <iustin@google.com>

Modify LUFailoverInstance to enable shared storage instances to failover.
Shared storage instance failover requires either a target node or an
iallocator to determine the target node. If none is given, the cluster default
iallocator is used.

The hook environment variables {OLD,NEW}_SECONDARY will be blank for shared
storage instances.

Locking behaviour is the same as for instance migration.

Signed-off-by: Apollon Oikonomopoulos <apollon@noc.grnet.gr>
[iustin@google.com: revert the DTS_NET_MIRROR specific changes]
Signed-off-by: Iustin Pop <iustin@google.com>
Reviewed-by: Iustin Pop <iustin@google.com>

Modify LUNodeMigrate to provide node migration for nodes with instances using
shared storage. gnt-node migrate has to be passed an iallocator for migration
of shared storage instances to be performed. When using a shared storage
backend, all cluster nodes are locked.

Signed-off-by: Apollon Oikonomopoulos <apollon@noc.grnet.gr>
Signed-off-by: Iustin Pop <iustin@google.com>
Reviewed-by: Iustin Pop <iustin@google.com>

Modify LUMigrateInstance and TLMigrateInstance to allow instance migrations for
instances with DTS_EXT_MIRROR disk templates.

Migrations of shared storage instances require either a target node, or an
iallocator to determine the target node. If none is given, the cluster default
iallocator is used.

Locking behaviour: If the iallocator is used, then initially all nodes are
locked and subsequently only the locks on the source node and the target node
selected by the iallocator are retained.

Signed-off-by: Apollon Oikonomopoulos <apollon@noc.grnet.gr>
[iustin@google.com: small changes in cmdlib.py]
Signed-off-by: Iustin Pop <iustin@google.com>
Reviewed-by: Iustin Pop <iustin@google.com>

Add DST_NODE_OPT to cli.py to use for directly specifying the target node
during migration/failover.

gnt-instance failover/migrate also get passed an iallocator option.

gnt-node failover/migrate get only a target_node option.

Signed-off-by: Apollon Oikonomopoulos <apollon@noc.grnet.gr>
Signed-off-by: Iustin Pop <iustin@google.com>
Reviewed-by: Iustin Pop <iustin@google.com>

Add iallocator and target_node slots to OpMigrateInstance and
OpFailoverInstance to facilitate shared-storage-backed instance mobility. Add
iallocator slot to OpMigrateNode (no explicit target_node in this case).

Signed-off-by: Apollon Oikonomopoulos <apollon@noc.grnet.gr>
Signed-off-by: Iustin Pop <iustin@google.com>
Reviewed-by: Iustin Pop <iustin@google.com>

Make cmdlib.IAllocator shared-storage-aware. IAllocator requires secondary
nodes only on DTS_NET_MIRROR disk templates and requires no secondaries for
DTS_EXT_MIRROR templates.

Signed-off-by: Apollon Oikonomopoulos <apollon@noc.grnet.gr>
Signed-off-by: Iustin Pop <iustin@google.com>
Reviewed-by: Iustin Pop <iustin@google.com>

This patch introduces basic shared block storage support.

It introduces a new storage backend, bdev.PersistentBlockDevice, to
use as a backend for shared block storage. The new bdev requires a new
BLOCKDEV_DRIVER_MANUAL constant with the value "manual" and uses it as
the first part of the block device unique_id.

A new disk template, DT_BLOCK is introduced as well and added to
DTS_EXT_MIRROR and DTS_MAY_ADOPT. Also added DTS_MUST_ADOPT constant
and use it to check for the presence of the adopt keyword during LU
invocation. We enforce the /dev/disk limitation upon adoption, but we
allow block devices to reside anywhere under /dev.

This is very basic support and includes no storage manipulation (provisioning,
resizing, renaming) which will have to be implemented through a "driver"
framework.

Signed-off-by: Apollon Oikonomopoulos <apollon@noc.grnet.gr>
[iustin@google.com: slight changes to bdev.py]
Signed-off-by: Iustin Pop <iustin@google.com>
Reviewed-by: Iustin Pop <iustin@google.com>

The bdev_sizes multi-node RPC call returns the sizes of the requested
block devices on the desired nodes. Its intended use is to verify the
existence of a block device on a given node for shared block storage
support.

Block device paths are expected to lie under constants.BLOCKDEV_DIR
("/dev/disk" by default), where persistent symlinks for block devices
are assumed to exist.

Signed-off-by: Apollon Oikonomopoulos <apollon@noc.grnet.gr>
[iustin@google.com: small changes in backend.py]
Signed-off-by: Iustin Pop <iustin@google.com>
Reviewed-by: Iustin Pop <iustin@google.com>

Use DTS_MIRRORED to indicate mirrored disk templates that allow
migrations/failover.

DTS_MIRRORED is the union of DTS_EXT_MIRROR and DTS_NET_MIRROR.

Signed-off-by: Apollon Oikonomopoulos <apollon@noc.grnet.gr>
Reviewed-by: Michael Hanselmann <hansmi@google.com>
Signed-off-by: Iustin Pop <iustin@google.com>
Reviewed-by: Iustin Pop <iustin@google.com>

Add shared file storage handling during cluster initialization.

Signed-off-by: Apollon Oikonomopoulos <apollon@noc.grnet.gr>
Signed-off-by: Iustin Pop <iustin@google.com>
Reviewed-by: Iustin Pop <iustin@google.com>

This patch introduces core file storage support, consisting of the following:

A configure-time switch for enabling/disabling shared file storage
support and controlling the shared file storage location:
--with-shared-file-storage-dir=.  Shared file storage configuration is then
available as _autoconf.ENABLE_SHARED_FILE_STORAGE and
_autoconf.SHARED_FILE_STORAGE_DIR and there is a cluster-wide ssconf
key named "shared_file_storage_dir" for changing the file location.

A new disk template named "sharedfile" (DT_SHARED_FILE), using
ganeti.bdev.FileStorage.

Auxiliary functions in lib/config.py to handle shared file storage.

Signed-off-by: Apollon Oikonomopoulos <apollon@noc.grnet.gr>
[iustin@google.com: small style fixes]
Signed-off-by: Iustin Pop <iustin@google.com>
Reviewed-by: Iustin Pop <iustin@google.com>

Add doc/design-shared-storage.rst to document the proposed changes and update
Makefile.am respectively.

Signed-off-by: Apollon Oikonomopoulos <apollon@noc.grnet.gr>
Signed-off-by: Iustin Pop <iustin@google.com>
Reviewed-by: Iustin Pop <iustin@google.com>

Set the default stderr logging level to WARNING so the relevant output
can be seen.

Signed-off-by: Stephen Shirley <diamond@google.com>
Reviewed-by: René Nussbaumer <rn@google.com>

This isn't only used during cluster merge.

Signed-off-by: Stephen Shirley <diamond@google.com>
Reviewed-by: Michael Hanselmann <hansmi@google.com>

There is currently no way to reset oob_program back to its default from
the cmdline, which causes problems for cluster-merge. This patch means
that the following now works:
  gnt-cluster modify --node-parameters oob_program=

Signed-off-by: Stephen Shirley <diamond@google.com>
Reviewed-by: Guido Trotter <ultrotter@google.com>

Nodes can return unknown instances, so we shouldn't use the name as an
index without checking.

Signed-off-by: Iustin Pop <iustin@google.com>
Reviewed-by: Michael Hanselmann <hansmi@google.com>

If opening the log file fails, then we shouldn't attempt to use that
variable.

Signed-off-by: Iustin Pop <iustin@google.com>
Reviewed-by: Michael Hanselmann <hansmi@google.com>

Signed-off-by: Iustin Pop <iustin@google.com>
Reviewed-by: René Nussbaumer <rn@google.com>

The cfgupgrade tool was designed to be idempotent, that means it could
be run several times and still give produce the correct result. Ganeti
2.4 moved the file containing the RAPI users to a separate directory
(…/lib/ganeti/rapi/users). If it exists, cfgupgrade would automatically
move an existing file from …/lib/ganeti/rapi_users and replace it with a
symlink.

Unfortunately one of the checks for this was incorrect and, when run
multiple times, replaces the users file at the new location with a
symlink created during a previous run.

In addition the “--dry-run” parameter to cfgupgrade was not respected.
Unittests are updated for all these cases.

Signed-off-by: Michael Hanselmann <hansmi@google.com>
Reviewed-by: Iustin Pop <iustin@google.com>

NEWS update and version bump.

Signed-off-by: Iustin Pop <iustin@google.com>
Reviewed-by: Michael Hanselmann <hansmi@google.com>

* devel-2.3:
  Fix LUClusterRepairDiskSizes and rpc result usage
  Fix RPC mismatch in blockdev_getsize[s]
  RAPI: fix evacuate node resource

Signed-off-by: Iustin Pop <iustin@google.com>
Reviewed-by: René Nussbaumer <rn@google.com>

Also specifies the comma-escaping feature.

Signed-off-by: Iustin Pop <iustin@google.com>
Reviewed-by: René Nussbaumer <rn@google.com>

* devel-2.2:
  Fix LUClusterRepairDiskSizes and rpc result usage
  Fix RPC mismatch in blockdev_getsize[s]

Signed-off-by: Iustin Pop <iustin@google.com>
Reviewed-by: René Nussbaumer <rn@google.com>

This LU was introduced before the RPC result conversion from .data to
.payload, and it has managed to keep the old-style usage (how? it's
the only LU that does so). Fix by changing to payload, and add some
extra logging for easier diagnose.

Signed-off-by: Iustin Pop <iustin@google.com>
Reviewed-by: Stephen Shirley <diamond@google.com>
Reviewed-by: Michael Hanselmann <hansmi@google.com>
(cherry picked from commit 043beb38)

Commit 92fd2250 added consistency checks in the RPC layer, which broke
the call_blockdev_getsizes RPC call (declared with 's' at the end in
rpc.py, without 's' in the node daemon).

The immediate fix is to correct the rpc function name, the long term
one will be to remove this duplication.

Signed-off-by: Iustin Pop <iustin@google.com>
Reviewed-by: Stephen Shirley <diamond@google.com>
(cherry picked from commit ccfbbd2d)

PollJob returns the whole op_results, hence a list of opcode results.

Signed-off-by: Iustin Pop <iustin@google.com>
Reviewed-by: René Nussbaumer <rn@google.com>

* origin/stable-2.4:
  Fix typo in kvm-ifup script
  NEWS: Replace smartquotes, start lines with uppercase
  Update NEWS and release 2.4.0 rc3
  Fix potential data-loss bug in disk wipe routines

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

Reported-by: Bas Tichelaar <bas@30loops.net>
Signed-off-by: Michael Hanselmann <hansmi@google.com>
Reviewed-by: Guido Trotter <ultrotter@google.com>