Age | Commit message (Collapse) | Author | Files | Lines |
|
Signed-off-by: Corentin LABBE <clabbe@baylibre.com>
|
|
If we use persistend_db, a restart with an image containing a backup
will re-apply the backup and so overrun current DB.
So when backup is used, apply a flag to not reuse it.
|
|
By setting pg_lava_password, the final lavaserver password will be set
to this value.
|
|
device-types are now in /usr/share/lava-server/device-types/ and so
device-type-patch does not work anymore.
This patch support both location
|
|
LAVA has changed the place of DB password from instance.conf to a yaml
file.
|
|
|
|
Pr 2020.01
|
|
This patch permits to handle migrations when upgrading.
|
|
When backup size grows to several GB, tar operation can fail
with "no space left on device" error.
This is because of the limited size of /tmp on certain machines
This changes allows using a different folder.
This can be done by simply adding such line in Dockerfile:
ENV TMPDIR /wherever/tmp
|
|
This patch backport a LAVA patch which fix an issue with lavacli
commands.
|
|
|
|
|
|
This patch upgrade LAVA to 2019.07 via their official docker images.
Along with the change of the baseimage from our lava-xx-base to official
2019.07, some minor changes are needed:
- Activate the en_US.UTF-8 locale needed for postgresql
- chown to lavaserver all copied device-types
- Fix the start scripts for using the official entrypoints
|
|
This patch permits to copy configurations files in lava-master/apache2/,
there will be copied over in the /etc/apache2 directory of master.
|
|
The lava-docker infrastructure have now splitted master and slave.
So there are no reason to keep tftp/lava-slave on master.
|
|
The start.sh will now be stored in lava-master.
This will help for migrating to official LAVA docker images
|
|
This patch adds an optional container which host healthchecks images.
|
|
Like for health-checks, device-types must be owned by lavaserver
|
|
Lava patch which contains device-type changes are ignored since final
device-types are already in /etc/lava-server/device-types/
So for fixing this, we apply lava-patch before device-type patch and
synchronize them.
|
|
|
|
Since 2019.03 is ready, use it
|
|
The current way to add patchs is non-determinist and could lead to
failure to build.
This patch made the order determinist by using sort
|
|
By default, all process use the DEBUG loglevel, this could be too much
for a production server.
This patch permits to tune the loglevel.
|
|
lava-server manage users list does not display inactive users.
Furthermore the return code of this action is not tested.
lava-server manage users list display also firstname/lastname and this
need to be filtered.
This patch fix thoses problem.
|
|
This patch made lava-docker use a known to work dockerhub tag.
Keep the latest tag for internal work.
|
|
The current way to do backup assume that slave will be restarted after
each master maintainance.
If master is restarted from backup without slaves restarting bring to
devices files missing.
By backuping devices, a master can be restored from backup without
needing slaves worker to be restarted.
|
|
For storing backups to be restored, it is better to use /root/backup/
instead of /
|
|
When there are no group, the grep return non-0 and setup.sh exits.
This patch fix by doing the filter in two step.
Fixes: e489f60b449b ("lava-master: skip creation of already existing groups")
|
|
When restoring a backup, it is not possible to create a group which
already exists.
This patch adds detection for this case and skip the creation when the
group exists.
|
|
This patch permits to add some env settings on slave.
The primary goal of this is to permits to add a proxy for a specific slave.
|
|
This fix allows users to add healthcheck files via API by changing
ownership from root to lavaserver. Without it users encounter a
permissions error.
Signed-off-by: Patryk Mungai <patryk.mungai-ndungu.kx@renesas.com>
|
|
This patch adds support for creating groups
|
|
This patch add the missing pyyaml requirement in the documentation,
fixing issue #31 in the process.
Note that this patch adds also a requirements.txt for easy pip install.
This patch also a note on http_fqdn stating this option as necessary
when using https.
|
|
|
|
|
|
The process of building LAVA is hard to maintain.
But since we need to be able to go back in time
(or just stay longer with a specific version)
we need a way to keep a LAVA version.
For achieving this, we build a minimal image with just LAVA
and tag it with the version of LAVA inside.
|
|
|
|
This patch add support for using ZMQ auth.
Basicly adding "zmq_auth: True" to a master is sufficient to enable it.
Since "ZMQ certificates" are using a custom format (vs X509 classic), we need to use the custom generator.
For helping with that a temporary docker is generated which handle generating thoses files.
|
|
When restoring a database, token are restored and so cannot be re-created.
|
|
This patch add an easy way to backup postgresql database and needed data stored in filesystem.
In the mean time, a way to restore backup during start is added
|
|
Until now, each docker-compose down will loose all data (jobs, devices, etc...).
By creating a docker volume, the database will now be persistent other restart.
Due to a risk of the lava package reseting the password, it is always reseted at start.
This patch add also another volume for job output which are stored in the filesystem and not in database.
|
|
|
|
This patch remove the use of setup.sh during build.
So setup.sh is now ran during startup.
Since setup.sh could ran multiple times (start -> stop -> start) against
the same database, we need to handle already existing objects (user,
devices,...)
So all "add objects" are done only after a check that this object is not
already present.
|
|
If backporting LAVA patch is necessary, this patch add an easy way to do
it.
|
|
It is unnecessary to copy start/stop twice during build
|
|
This patch permits to access LAVA webadmin via HTTPS.
Resolve SPEC-1493
|
|
|
|
This patch upgrade lava-docker to LAVA 2018.4
|
|
This patch implement multiple slave support.
Instead of having a maximum of one master and one slave in one docker
image, it is now possible to have multiple slave accross several docker
host.
For helping this change, a new boards.yaml format is introduced (See
README.md for details)
Note that tokens.yaml is also squashed in boards.yaml
|
|
qemu-kvm is not availlable on ARM.
Install it only on supported x86/x86_64
|