Statistics
| Branch: | Tag: | Revision:

root / src / Ganeti @ c3a70209

Name Size
  Confd
  Cpu
  Curl
  DataCollectors
  HTools
  Hs2Py
  Hypervisor
  Monitoring
  Query
  Storage
BasicTypes.hs 8.1 kB
Common.hs 14.4 kB
Compat.hs 1.4 kB
Config.hs 14.8 kB
ConfigReader.hs 11.7 kB
ConstantUtils.hs 5.4 kB
Constants.hs 113.9 kB
Daemon.hs 15.6 kB
Errors.hs 6.4 kB
Hash.hs 1.8 kB
JQueue.hs 15.9 kB
JSON.hs 10.3 kB
Jobs.hs 3.3 kB
Logging.hs 4.6 kB
Luxi.hs 16.2 kB
Network.hs 3.5 kB
Objects.hs 24.4 kB
OpCodes.hs 26.7 kB
OpParams.hs 45.6 kB
Parsers.hs 1.7 kB
Path.hs 3.6 kB
PyValueInstances.hs 2.4 kB
Rpc.hs 19.3 kB
Runtime.hs 7.1 kB
Ssconf.hs 5.1 kB
THH.hs 46.9 kB
Types.hs 27.7 kB
Utils.hs 18.4 kB
Version.hs.in 246 Bytes

Latest revisions

# Date Author Comment
c3a70209 11/28/2013 04:59 pm Klaus Aehlig

Document the jobqueue timestamp format

...and also provide a method to get the current time in
that format.

Signed-off-by: Klaus Aehlig <>
Reviewed-by: Helga Velroyen <>
fa7f3e26 11/27/2013 03:08 pm Helga Velroyen

Fix removal of duplicates

Commit ede6df3d02 introduced a bug in the node queries
where disk templates where paired up wrongly to their
storage unit keys due to removal of duplicates at the
wrong place. This patch fixes it.

Signed-off-by: Helga Velroyen <>...
96e3dfa7 11/27/2013 11:37 am Helga Velroyen

Fix retrieval of number of instances of a node

This patch fixes a FIXME to make the retrieval of the
number of primary and secondary instances share more
common code.

Signed-off-by: Helga Velroyen <>
Reviewed-by: Klaus Aehlig <>
e86c9deb 11/27/2013 11:37 am Helga Velroyen

Use hypervisor / storage information only when requested

So far, the node queries ignored the list of fields and
just requested all available information from the backend.
That means, for example if only hypervisor information is
requested, still the storage space calculation is...
ede6df3d 11/27/2013 11:37 am Helga Velroyen

Remove duplicate storage units in node query

This is a little performance tweak for the node queries.
So far, the query code mapped the disk templates to storage
units. It could happen that two disk templates were mapped
to the same storage unit and therefore the storage space...
f5b765f0 11/26/2013 03:31 pm Klaus Aehlig

Make luxid job submission be defined by replication

When receiving jobs to be submitted, make luxid replicate them to all
master candidates and then return. The actual execution can be handled
asynchronously.

Signed-off-by: Klaus Aehlig <>...
493d6920 11/26/2013 03:31 pm Klaus Aehlig

Add function to enqueue jobs

Add a function that ensures that a given set of jobs gets executed at
the appropriate time. At the moment, this is still the simple
mechanism of handing over everything to masterd; but even at this
stage, it has the benefit of allowing to remove code duplication in...
ea128e20 11/26/2013 03:31 pm Klaus Aehlig

Add a function justBad to filter the Bad value of a list

In the same way as justOk allows to filter the Ok values,
add justBad to filter the Bad values. While there, simplify
the definition of justOk.

Signed-off-by: Klaus Aehlig <>
Reviewed-by: Petr Pudlak <>
9fd653a4 11/26/2013 03:31 pm Klaus Aehlig

Add wrapper to replicate many jobs

Add a convenience wrapper around replicateJob to replicate
many jobs to the master candidates.

Signed-off-by: Klaus Aehlig <>
Reviewed-by: Petr Pudlak <>
b5a96995 11/26/2013 03:31 pm Klaus Aehlig

Add function to replicate a job to the master candidates

As luxid will be handling the job queue soon, add a utility to
replicate jobs to all master candidates. Also log errors.

Signed-off-by: Klaus Aehlig <>
Reviewed-by: Petr Pudlak <>

View revisions

Also available in: Atom