- moved hosts parameter from mandatory to optional field in CreateBlockVolume
method of BlockProvider interface,since hosts field may not required for
other block providers like loopback.
- a common function for updating available hosting volume size will prevent
from duplicate code
Signed-off-by: Oshank Kumar <okumar@redhat.com>
Changes have been made to enabled them from the config file.
Setting "profiling = true" in config file will enable it.
The bottleneck can be understood by seen the output using
the go tool pprof.
example: go tool pprof http://<ip>:24007/debug/pprof/heap
Signed-off-by: Hari Gowtham hgowtham@redhat.com
Category prefix is optional when setting Volume options. For example
to set replicate.eager-lock, we can pass `replicate.eager-lock` or
`cluster/replicate.eager-lock`. With this PR, always stored with
category prefix.
Also fixed the issue of loosing template variables when xlator default
options and volinfo.Options are loaded(Fixes: #1397)
Signed-off-by: Aravinda VK <avishwan@redhat.com>
Currently client Volfiles are stored in etcd, but this increases the
consumption of etcd storage space. With this PR, client volfiles are
generated on the fly when mount request comes.
Signed-off-by: Aravinda VK <avishwan@redhat.com>
Add a section in the "Troubleshooting and Debugging" section of the
glusterd2 development guide on how to setup and view traces for quick
debugging.
Signed-off-by: Sridhar Seshasayee <sseshasa@redhat.com>
Brick volfiles generation added during Snapshot activate and
volfiles delete added during snapshot deactivate.
Also fixed client volfile regeneration of snapshot volume during
snapshot restore undo.
Signed-off-by: Aravinda VK <avishwan@redhat.com>
Correction for fixing the request and response for Methods: GeoReplicationConfigGet and GeoReplicationConfigSet. Fixed the code in plugins/georeplication/init.go. Also regenerated the endpoints.md file.
Signed-off-by: Sidharth Anupkrishnan <sanupkri@redhat.com>
With respect to https://github.com/gluster/glusterd2/pull/1123. Using golang 1.9 or higher instead of 1.8. @prashanthpai Please check if we need to make changes in any other documentation.
Add some opinions and conventions about:
* logging of txn.* methods
* logging using WithError()
* use of goto statements
* add links pointing to go proverbs and good commit messages
Signed-off-by: Prashanth Pai <ppai@redhat.com>
Implemented wehook/test rest API in glusterd2
which will be called by clients to test webhook
connections between all peers in glusterd2
and webhook URL
added rest client code to call webhook/test API
Signed-off-by: Madhu Rajanna <mrajanna@redhat.com>
For all practical purposes, in real usage and in tests, the configurable
paths `workdir` and `localstatedir` are set to the same directory which
defaults to `/var/lib/glusterd2` or `/user/local/var/lib/glusterd2`
Unlike glusterd1, glusterd2 does not daemonize itself. The working
directory of the process can be meaningfully set to either current
working directory from where glusterd2 process is started or to
`localstatedir` where glusterd2 stores all its data. We choose the
latter.
This change also reduces ambiguity for users and one less thing to
configure.
Signed-off-by: Prashanth Pai <ppai@redhat.com>
consumers wanted to
to check the status of glusterd2.
they can consume this URL to check
the status of glusterd2.
Signed-off-by: Madhu Rajanna <mrajanna@redhat.com>
