This patch changes the multi-instance gnt-* commands (gnt-instance
start/stop, gnt-node evacuate/failover) such that the individual
operations are submitted in parallel, ideally improving the speed of the
execution.

The patch does this by abstracting the job set functionality into a new
class in cli.py, that takes care of the job submit, job poll and error
handling.

Reviewed-by: ultrotter

This patch makes gnt-cluster copyfile and command skip the offline
nodes.

Reviwed-by: ultrotter, imsnah

A job queue with too many jobs can increase memory usage and/or make
the master daemon slow. The current limit is just an arbitrary number.
A "soft" limit for automatic job archival is prepared.

Reviewed-by: iustinp

This patch should fix all outstanding epydoc parsing errors; as such, we
switch epydoc into verbose mode so that any new errors will be visible.

Reviewed-by: imsnah

- Add new "canceling" status
- Notify clients when job is canceled
- Give a return value from CancelJob
- Handle it in the client library

Reviewed-by: iustinp

Currently, job failures are done by raising OpExecError(job result). For
a one-opcode job that failed, this is very non-intuitive:
Failure: command execution error:
[u'Disk size change not possible, use grow-disk']

This patch changes the output in two ways:
 - for a job which didn't have opcodes which succeeded, it changes it
   to:
   Failure: command execution error:
   Disk size change not possible, use grow-disk
 - for a job with partial completion, to:
   Failure: command execution error:
   partial failure (opcode 1): Cannot remove the last disk of an
   instance

The actual error type is not changed. A job that doesn't have at least
an error-ed opcode is shown as before.

Reviewed-by: imsnah

This patch enables by default the old 'human-readable' option, but in a
slightly different model.

The option is now called "units" and takes either:
 - 'h' for automatic formatting
 - 'm', 'g' or 't' for mebi/gibi/tebibytes

If 'h' is used, we add a unit suffix, otherwise nothing is added so that
parsing is easy.

The default value of this unit is:
  - 'h' if a separator is not passed
  - 'm' if a separator is passed

Reviewed-by: ultrotter

This changes cli.py to FieldSet usage so that gnt-instance list will
format nicely the disk.size/*, and the count of disks/nics.

Reviewed-by: ultrotter

ganeti-masterd: Add initialization and shutdown of RPC pool. It needs
to be shutdown before forking.

ganeti.cli: Add decorator function to initialize and shutdown RPC pool.

ganeti.rpc: Add functions to initialize and shutdown RPC pool. Throw
exception when used without proper initialization.

gnt-cluster, gnt-node: Use decorator function to initialize and shutdown
RPC pool.

Reviewed-by: iustinp

This patch adds support for querying in gnt-instance list of:
  - disk.count
  - nic.count
  - disk.size/$N
  - nic.(ip|mac|bridge)/$N

The patch also disables the exception raised when the header description
is missing, that should be fixed in the future (together with better
handling of variable numeric fields).

Reviewed-by: imsnah

Currently we don't catch the JobSubmittedException (although the code to
format it nicely is there). The patch adds the exception name to the
catch list and makes --submit work.

Reviewed-by: imsnah

Since now we use only one function from the logger module
(SetupLogging), we move it to utils.py (which is already imported by all
users of this function), and we remove the module.

Reviewed-by: imsnah

We also add two function for printing messages, so that scripts won't
have to import logger to get these. They are a simple extension over the
logger ones, as they accept the call style from logging:
 ToStdout("Message: %s", msg)
(instead of requiring formatting by the client.

Reviewed-by: imsnah

We add a (per-node) queue drain flag that blocks new job submission.
There is not yet an interface to add/remove the flag (will come in next
patches).

Reviewed-by: imsnah

This is used to validate the parameters in both “gnt-instance add” and
“gnt-backup import”.

Reviewed-by: ultrotter

This patch adds paramter removal in SplitKeyVal, by prefixing a
value-less key with "-"; this is needed in resetting parameters back to
cluster defaults, but care must be applied now that None can come from
the parser.

Reviewed-by: imsnah

For the new 2.0-style command line options, we need to parse strings of
the type:
  ident:key=val[,...]
and
  key=val[,...]

This patch adds two new option builders for these two, which return
(ident, {key=val,}) and {key=val,} for the above two formats. It also
handles specially constructs of type “key” (val is set to True) and
“no_key” (val is set to False, and the ‘no_’ prefix is stripped).

Reviewed-by: imsnah

This patch adds a new luxi call that implements auto-archiving of jobs
older than a certain age (or -1 for all completed jobs), and the gnt-job
command that makes use of this (with 'all' for -1).

Reviewed-by: imsnah

This function will be used for auto-archiving jobs via the command line.
The function is pretty simple, we only support up to weeks since months
and higher are not 'precise' entities, and dealing with them would
require us to start using calendar functions.

Reviewed-by: imsnah

In case the job object doesn't have a timestamp (which is a separate
issue), the listing should not break. We fix this by changing the
FormatTimstamp function itself to return '?' in case the timestamp
doesn't look good (note that it still can break if non-integers are
returned, but this is unlikely).

Reviewed-by: imsnah

Curently PollJob accepts a generic job, but will return (history
artifact) only the first opcode result. This is wrong, as it doesn't
allow polling of a job with multiple results.

Its only caller (for now) is also changed, so no functional changes
should happen.

Reviewed-by: ultrotter, amishchenko

Currently we format the timestamp inside the gnt-job info function. We
will need this more times in the future, so move it to cli.py as a
separate, exported function.

Reviewed-by: imsnah

The "sys.exit(0)" was not nice as you couldn't differentiate it from
other exit codes. We change this to a specially defined exception for
this, so that multi-opcode commands can handle this nicely.

Reviewed-by: imsnah

Reported by Iustin.

Reviewed-by: iustinp

This is a large patch, but I can't figure out how to split it without
breaking stuff. The old way of getting messages by always getting the
last one didn't bring all messages to the client if they were added
too fast, thereby making commands like “gnt-cluster verify” less than
useful. These changes now introduce some sort a serial number per
log entry to keep track what message a client already received. They
also remove the log lock per opcode to make reading log entries thread
safe.

Reviewed-by: ultrotter

This means commands will not take at least one second anymore.

Reviewed-by: ultrotter

See the comment in the patch.

Reviewed-by: ultrotter

feedback_fn wasn't passed to it.

Reviewed-by: iustinp

This patch adds the infrastructure for executing a job in background,
instead of foreground, via a new “--submit” option. The behaviour is
that the job ID is printed and the script will immediately exit.

The patch also converts gnt-node list to this model (yes, this will be a
query in the future).

Reviewed-by: imsnah

The current SubmitOpCode function is not flexible enough to be used for
submitters that don't want to wait for the job finish.

The patch splits this in two, a SendJob function and a PollJob one, and
the old SubmitOpCode becomes a wrapper. Note that the new SendJob takes
a list of opcodes (and not a single opcode anymore).

Reviewed-by: imsnah

This patch adds a check that we are on the master after failing to
connect to the socket, and log nicely the master name.

Reviewed-by: ultrotter

This is mostly:
  - whitespace fix (space at EOL in some files, not all, broken
    indentation, etc)
  - variable names overriding others (one is a real bug in there)
  - too-long-lines
  - cleanup of most unused imports (not all)

Reviewed-by: ultrotter

The 'old-style' info, error, debug logs do not make much sense. This
patch unifies the SetupLogging and SetupDaemon functions. As a result,
all the commands logs to a 'commands.log' file.

The patch also changes the log setup to keep going if there's an error
in setting up the file logging but we're logging to stderr.

Also, burnin now logs to its own file (burnin.log).

Reviewed-by: ultrotter

Currently, the feedback_fn argument to SubmitOpCode is no longer used.
We still need it in burnin, so we re-enable it by making the code call
that function with the msg argument in case feedback_fn is callable. The
patch also modifies burnin to accept the new argument format (msg is not
a triple instead of a string).

The patch also removes the ‘proc’ argument as it's obsolete; instead we
can accept a luxi.Client instance (noone uses this right now).

Reviewed-by: ultrotter

This patch contains a raw version for fixing feedback_fn.

The new mechanism works as follows:
  - instead of a per-Processor feedback_fn, there's one for each
    ExecOpCode, so that feedback for different opcodes go via possibly
    different functions
  - each _QueuedOpCode gets a message buffer, a method for adding
    feedback and a method for retrieving (parts of) the feedback
  - the _QueuedJob object gets a new attribute that is equal to the
    index of the currently executing opcode
  - job queries get an extra parameter called 'ticker' that will return
    the latest message on the current executing opcode
  - the cli.py job completion poll will show the new status if different
    from the old one

Of course, quick messages will be lost, as currently only the latest one
is available. Also changes between opcodes are not represented at all.

Reviewed-by: imsnah

"gnt-job list" was broken after my recent changes in the RPC
between clients and the master. This patch makes it work again.

Reviewed-by: iustinp

- Introduce abstraction class on client side
- Use constants for method names
- Adopt legacy function SubmitOpCode to use it

Reviewed-by: iustinp

By a mistake, we don't reuse the luxi client. As such, we open and close
the connection at each poll cycle and spam the server logs.

Reviewed-by: ultrotter

Currently the generic handling of ganeti errors in cli.py (GenericMain
and FormatError) only handles the core ganeti errors, and not the client
protocol errors (which live in a separate hierarchy).

This patch adds handling of luxi errors too, and also adds another luxi
error for the case when the master is not running. This gives us a nice:

  gnta1:~# gnt-node list
  Cannot communicate with the master daemon.
  Is it running and listening on '/var/run/ganeti-master.sock'?

error message instead of a traceback.

Reviewed-by: amishchenko

- Shorter code
- Reorder arguments to logger.SetupLogging calls to make more sense

Reviewed-by: iustinp