This patch adds a new opcode and lu for checking disk sizes. Currently
it does only top-level disk verification, and also doesn't check
primary/secondary node size mismatches (these two are added as TODOs in
the Exec() function of the LU).

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

This patch modified OpActivateDisks, LUActivateDisks and gnt-instance
activate-disks to support and pass this option to
_AssembleInstanceDisks.

The patch is quite trivial I think; there should be no issues from it
except if used when not needed.

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

This is identical to dc30b0e4 but applied to gnt-backup. Thanks to user
ocaner for catching it.

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

This allows failing over in certain corner cases, such as a 2 node
cluster with one node down. The man page is also updated to document
this dangerous option and how to recover from this situation.

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

This patch fixes a few node readd issues.

Currently, the node readd consists of two opcodes:
  - OpSetNodeParms, which resets the offline/drained flags
  - OpAddNode (with readd=True), which reconfigures the node

The problem is that between these two, the configuration is inconsistent
for certain cluster configurations. Thus, this patch removes the first
opcode and modified the LUAddNode to deal with this case too.

The patch also modifies the computation of the intended master_candidate
status, and actually sets the readded node to master candidate if
needed. Previously, we didn't modify the existing node at all.

Finally, the patch modifies the bottom of the Exec() function for this
LU to:
  - trigger a node update, which in turn redistributes the ssconf files
    to all nodes (and thus the new node too)
  - if the new node is not a master candidate, then call the
    node_demote_from_mc RPC so that old master files are cleared

My testing shows this behaves correctly for various cases.

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

This patch adds a ‘role’ node list field, which shows a one-character
node status. This is a simpler way to see the node status than selecting
all the flags individually.

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

Currently running “gnt-instance list -o+vcpus” fails with a cryptic message:
  Unhandled Ganeti error: vcpus

This is due to multiple issues:
  - in some corner cases cmdlib.py raises an errors.ParameterError but
    this is not handled by cli.py
  - LUQueryInstances declares ‘vcpu’ as a supported field, but doesn't handle
    it, so instead of failing with unknown parameter, e.g.:
      Failure: prerequisites not met for this operation:
      Unknown output fields selected: vcpuscd
    it raises the ParameteError message

This patch:
  - adds handling of 'vcpus' to LUQueryInstances
  - adds handling of the ParameterError exception to cli.py
  - changes the 'else: raise errors.ParameterError' in the field handling of
    LUQueryInstance to an assert, since it's a programmer error if we reached
    this step

With this, a future unhandled parameter will show:
  gnt-instance list -o+vcpus
  Unhandled protocol error while talking to the master daemon:
  Caught exception: Declared but unhandled parameter 'vcpus'

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

The size of the instance's disk was not shown in “gnt-instance info”.
This patch adds it and formats it nicely if possible.

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

This patch modifies the start instance script, opcode and logical unit
to support temporary startup parameters.

Different from 1.2, where only the kernel arguments were supporting
changes (and thus xen-pvm specific), this version supports changing all
hypervisor and backend parameters (with appropriate checks).

This is much more flexible, and allows for example:
  - start with different, temporary kernel
  - start with different memory size

Note: in later versions, this should be extended to cover disk
parameters as well (e.g. start with drbd without flushes, start with
drbd in async mode, etc.).

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

Currently QueryJob returns "None" when a wrong job ID is passed.
Handle this in gnt-job list, by printing an error for each wrong job,
and still giving output for all the jobs which actually do exist.

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

Currently doing a gnt-cluster-modify --no-lvm-storage is silently
ignored, as it passes a None value in vg_name, which is the same as not
modifying that parameter. Explicitely set the passed value to '', so the
non-true not-None value can be evaluate to actually remove a volume
group.

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

Even if we cannot modify all of them, they are useful information about
the current cluster.

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

The _TransformPath function is not used anymore in 2.0, let's remove it.

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

Currently “gnt-debug submit-job” takes a single argument and has
non-trivial startup-costs; in order to exercise the job system, it is
better to be able to submit multiple jobs with a single invocation of
the script.

This patch extends it to take multiple argument, de-serialize the
opcodes and then submit all of them as fast as possible, in order to
increase pressure on the master daemon.

Signed-off-by: Iustin Pop <iustin@google.com>
Reviewed-by: Alexander Schreiber <als@google.com>

The current implementation of “gnt-cluster getmaster” doesn't work on
non-master nodes, which is a regression from 1.2. This patch implements
it (again) via ssconf.

Signed-off-by: Iustin Pop <iustin@google.com>
Reviewed-by: Alexander Schreiber <als@google.com>

Don't show all instances info by default, but require --all to be passed
for this time consuming operation.

Reviewed-by: iustinp

This patch adds a couple of both externally and internally reported
issues:
  - missing SGML tags (Issue 54), report and patch by superdupont
  - wrong variable used in the init.d script, report and patch by
    Karsten Keil <karsten-keil@t-online.de>
  - man page for gnt-instance reinstall needs clarification (Issue 56)
  - gnt-instance man page missing --disks documentation for
    replace-disks
  - gnt-node modify help output is unclear about the -C/-D/-O input
    format, and the man page doesn't document this command at all
  - “gnt-node modify -C yes” for offline or drained nodes had wrong
    error message
  - “gnt-instance reinstall --select-os” has wrong prompt, we only
    accept a number for the OS and not the template name

Reviewed-by: ultrotter

This patch raises an error in the master daemon in case the user
requests a locking query; accordingly, all clients were modified to send
only lockless queries. This is short-term fix, for proper fix the
clients should be modified to submit a job when the user request a
locking query.

The other approach would be to ignore the flag passed by the client;
this would be worse as client's wouldn't get at least an error.

The possible impact of this is multiple:
  - some commands could have been not converted, and thus fail; this
    can be remedied easily
  - the consistency of commands is lost; e.g. node failover will not
    lock the node *while we get the node info*, so we could miss some
    data; this is again in the thread of atomic operations which are
    missing in the current model of query-and-act from gnt-* scripts

Reviewed-by: imsnah, ultrotter

 # gnt-cluster queue foo
 Failure: prerequisites not met for this operation:
 Command 'foo' is not valid.

Reviewed-by: iustinp

Similar to the --disk fixes a while ago, --net is broken too. This patch
fixes it.

Reviewed-by: imsnah

This patch removes the extra_args parameter and instead switches the
instance to the HV_KERNEL_ARGS hypervisor option.

This is a big change, but it's a needed cleanup, this extra parameter on
all RPC calls is not generic and we also need to have a persistent value
here.

Reviewed-by: imsnah

Having hvattr descriptions is only confusing for the user, because even
if they explain better what an attribute is about, they don't help in
deciding what keyword should be used to actually set it. If in the
future we want descriptions they should probably live in constants.py,
and be displayed together with the key, rather than instead of it.

This patch also changes the handling of the vnc console connection
description, making compatible work with both kvm and xen-hvm.

Reviewed-by: iustinp

This patch adds back to the instance creation command (gnt-instace add,
gnt-backup import) the ‘-s’ short form option for specifying a
single-disk instance.

Also a small bug in gnt-backup import is fixed.

Reviewed-by: ultrotter

This patch changes the gnt-node and gnt-job list commands to accept
argument and list only the selected items, which is useful when having
many nodes or jobs.

It also removes the “--units” option from gnt-job list as we don't
actually use it.

Reviewed-by: imsnah

This patch changes the scripts so that the short name for the
“--iallocator” option is always ‘-I’.

Reviewed-by: ultrotter

Currently the batcher hypervisor parameter must be a dict with one
element (e.g. {"xen-hvm": { "acpi": true }}). This is overly complex and
hard to validate correctly; the patch splits it in two:
  - one "hypervisor" string parameter, with the name of the hypervisor
  - one "hvparams" dictionary, with the hypervisor parameters

The patch also changes the error handling in parsing the definition file
- since this is not a long-running file, we are less concerned with safe
closing of the file, and more with presenting meaningful error
messages.

Reviewed-by: killerfoxi

It's hvparams, not opts.hvparams.

Reviewed-by: iustinp

If hvparams is not set it will be [], so dict() will transform it to an
empty dict, which is safe in all cases.

Reviewed-by: iustinp

The patch sorts the instance list in gnt-node info output, in order to
make it more readable (and stable).

Reviewed-by: imsnah

The patch changes the pre-checks in node-add and re-add:
  - if the node is not already in the cluster, refuse to re-add
  - when re-adding, reuse the secondary IP from the cluster
    configuration
  - when re-adding, reset the offline and drained flags, so that RPC
    calls work (and we can actually upload the keys)

The patch also adds a missing log entry in LUSetNodeParams.

Reviewed-by: imsnah

We want all the hv/be parameters to have a known type, rather than a
random mix of empty string, boolean values, and None, so we declare the
type of each variable and we enforce/convert it.

- Add some new constants for enforceable value types
- Add new constants dicts HVS_PARAMETER_TYPES and BES_PARAMETER_TYPES
  holding not only the valid parameters but also their types
- Drop the old HVS_PARAMETERS and BES_PARAMETERS constants and calculate
  the values from the type dict
- Convert all the default parameters to a valid type value
- Create a new ForceDictType utils function, to check/enforce a dict's
  element value types, with relevant unit tests
- Drop a few custom functions to check/convert the BE param types in
  utils and cli, in favor of ForceDictType
- Double-check the parameter types using ForceDictType in both scripts
  and LogicalUnits, when possible.

As a bonus:
- Remove some old commented-out code in gnt-instance
- Remove some already fixed FIXME
- Fix a bug which prevented VALUE_DEFAULT to be applied to BE parameters
  in SetInstanceParams because the value was checked for validity before
  that transformation was made
- Fix a bug which prevented initing a cluster and passing hvparams to
  work at all
- ForceDictType allows an allowed_values for exceptions, which makes us
  able to do the checking even when some values must not be
  converted/typechecked (for example the 'default' string in
  SetInstanceParameters)

Reviewed-by: iustinp

This patch adds LU and cli-level support for modification of the node
drained flag. It is similar to the offline changes.

Reviewed-by: imsnah

This patch exports the drained attribute:
  - LUQueryNodes accepts now the drained field
  - RAPI exports it for node objects
  - gnt-node info shows it now (along newly-added master_candidate and
    offline flags)
  - gnt-node list can list it (but not by default)
  - to the iallocator scripts

Reviewed-by: imsnah

This patch adds a new instance query flag called disk_usage that
retrieves the overall space used by an instance on each of its nodes.
This can be used when balancing the cluster or checking N+1 status.

The flag is also exported in RAPI. Note the flag is currently broken for
file-based instances, as it represents the amount of space in the
cluster volume group.

Reviewed-by: ultrotter

This is a hand-picked forward patch of commit 1755 on the 1.2 branch
(hand-picked since the trees diverged too much since then):

    The patch changed the xen hypervisor to compute the number of cpu
    sockets/nodes and enables the command line and the RAPI to show this
    information (for RAPI is enabled by default in node details, for gnt-one
    one can use the new “cnodes” and “csockets” fields).

    Originally-Reviewed-by: ultrotter

For the KVM and fake hypervisors, the patch just exports 1 for both
nodes and sockets. This can be fixed, by looking at the
/sys/devices/system/cpu/cpuN/topology directories, and computing the
actual information, but that should be done in a separate patch.

Reviewed-by: imsnah

This patch converts some more jobs with only queries into cheaper luxi
queries (no job created), and fixes some fallout from the lockless
queries changes.

Reviewed-by: ultrotter

Similar to the instance list, this patch enables lockless node queris.
“gnt-node list” accepts now the “--sync” flag which enables locking, the
default is lockless.

Reviewed-by: imsnah

This patch adds the online node list and instance list to the ssconf
keys. In order to do distribute correctly the instance list, we need to
update the cluster serial number on instance additions and removals.

The patch also changes the permissions on the ssconf files to be 0444:
  - no write for root, in order to signal that these file should not be
    modified
  - read for everyone since the files don't contain sensitive data
    anymore (and permissions can be controlled via the parent directory
    if needed)

The patch also fixes a small typo on gnt-cluster.

Reviewed-by: ultrotter

This patch adds the framework for, and enables lockless OpQueryInstances. This
means that instances will be shown in ERROR_up or ERROR_down state, even though
this is not an error (but just an in-progress job).

The framework is implemented as follows:
  - the OpQueryInstances, OpQueryNodes and OpQueryExports opcodes take
    an additional “use_locking” flag which will denote whether to lock
    or not; this patch only implements this for LUQueryInstances
  - the luxi query functions take an additional argument use_locking
    which is passed to the master daemon, and then passed to the above
    opcodes
  - cli.py export a new SYNC_OPT command line options which implement
    setting this flag to true
  - except for gnt-instance list, which uses this option, and for
    name-only queries (e.g. QueryNodes(fields=["names"])), all other
    callers are setting this flag to True
  - RAPI also sets the flag to True

The patch was tested with a continuous (0.2s sleep in-between)
gnt-instance list during a burnin, and no problems were observed.

Reviewed-by: ultrotter

This is a partial implementation of fully automated node evacuation:
we allow passing an iallocator and all instance replace-disks will be
execute via that iallocator.

The individual OpReplaceDisks opcodes are submitted in a single job,
which causes them to be executed serially and thus keeps the iallocator
runs consistent. This also changes the behaviour so that the first
reallocation that failed will stop all the reallocations.

Reviewed-by: ultrotter