* [bgp] Save admin state and set default state to shutdown
* Set default behavior to no shutdown
* Add build option SHUTDOWN_BGP_ON_START
* Script change for default admin state to be on
* Address CR comments to bgp_neighbor script
* Fix script bug
* Bump sonic-platform-modules-arista submodule
* Use sonic_sfputil plugin from the arista library
* Fix undefined variable varlog_size
* Prevent minigraph.xml to be removed from the flash
* Update DCS-7050QX-32 sensors config
If routes are inserted in main table, they wont' be overwritten
by the routes learned via BGP. Then the routes will be missing
from the ASIC.
Thus a default table is used so that control plane and data plane
routes are isolated and could be configured independently.
- Now that logrotate is a cron job that runs every minute, it was polluting syslog
- Also shrink max size of less-important logs to 50MB and rotate them daily by default
* [rc.local]: Copy saved minigraph if available.
In case of sonic-to-sonic update old image stores minigraph under /host
directory. Upon first boot this minigraph will be used by new image to
save configuration.
Signed-off-by: marian-pritsak <marianp@mellanox.com>
- rsyslog logs were being rotated regardless of whether they exceeded their maixmum size. This was due to "-f" flag passed to logrotate in cron job.
- After rotation, /var/log/syslog was never written to again. Instead, logs were written to /var/log/syslog.1. This was due to rsyslog not properly closing the file descriptor to the pre-rotated log.
- Also brought back time-related rotation via the new(er) maxsize option, which performs a boolean OR operation. If the log exceeds the maxsize OR the log hasn't been rotated in the specified, it will be rotated. Using the older size option, the time-based rotation was ignored.
- Also addresses issue #528
* Rename 'ACSFileFormat' -> 'SONiCFileFormat'
* Rename '00-acs.conf' -> '00-sonic.conf'
* Add logrotate.d and systemd-journald config files to image
* Log all SONiC process messages to /var/log/syslog; prevent duplicate logging to /var/log/messages
* Do not redirect cron and daemon logs to their own files, let them log to /var/log/syslog
* Log all teamd messages to /var/log/teamd.log; Add more SONiC program names to SONiC rules clause
* Remove duplicate code by condensing quagga programs into a list; Fix teamd log rule
* Kernel and LPR messages no longer getting duplicated to their own log files
* Now calling logrotate every minute via cron job
* Need full path to logrotate in cron job
* Add '.log' suffix to wildcards, otherwise logrotate will rotate already-rotated logs (e.g., bgpd.log.1.1.1.1.1...)
* Add microsecond granularity to syslog messages
* Don't overwrite system crontab, instead, install additional logrotate crontab file into /etc/cron.d
* Removed incomplete concept of per-process SONiC logs. We can revisit again later
* [device]: Add support for Mellanox MSN2410
MSN2410 runs on Spectrum silicon and has 56 ports:
48 25GbE and 8 100GbE
* Avoid full path within bash -c
Signed-off-by: marian-pritsak <marianp@mellanox.com>
- Do not bring up LAG member ports when LAG is not created.
This is because LAG member ports must be DOWN when joining
the teamd instances due to teamd design. Therefore, we cannot
bring up a LAG member port first and then join the port to a LAG.
Signed-off-by: Shuotian Cheng <shuche@microsoft.com>
* [build]: Include SONiC version into installer.
Signed-off-by: marian-pritsak <marianp@mellanox.com>
* Append dirty if contains local changes
Signed-off-by: marian-pritsak <marianp@mellanox.com>
* Update config
* Use correct name for kernel version field
* Update sysDescription.j2
- This change is added so that when doing ifdown/ifup to flap the
interface, it will success even when the interface is already
enslaved as a member of VLAN.
- LAG members do not have IP addresses. Thus ifconfig command is
needed.
- Do not display comments when not necessary.
Signed-off-by: Shuotian Cheng <shuche@microsoft.com>
* [base image]: Install SwSS SDK Python packages (Python 2 & 3) into base image
* Fix typo in existing variable name
* Only install Python 2 version, as Python 3 is not installed in base image
* [Aboot]:remove no need files before extract image
Signed-off-by: Sihui Han <sihan@microsoft.com>
* use and remove boot0 in the check
Signed-off-by: Sihui Han <sihan@microsoft.com>
* change the comments accordingly
Signed-off-by: Sihui Han <sihan@microsoft.com>
- This change ensures that when shuting down, opennsl-modules will
not stop until swss stops. Otherwise it will cause kernel stack
trace.
Signed-off-by: Shuotian Cheng <shuche@microsoft.com>
- Create /var/run/redis/ folder on the host
- Install Python client for Redis on the host
- Mount /var/run/redis/ as read/write from host for all dockers
- Enable accessing the database everywhere including on the host and from remote
Signed-off-by: Shuotian Cheng <shuche@microsoft.com>
When Type=simple, systemd will consider the service activated immediately
after specified in ExecStart process is started. If there is downstream
service depending on the state prepared in ExecStart, there will be race condition.
For example, issue #390. In this case, database.service calls database.sh, which
calls docker run or docker start -a to start database container. However, systemd
considers database.service successfully started at the time database.sh begins,
not after docker run finishes. As database.service is consider started, bgp.service
can be started. The redis database, which bgp service depends on, might or might not
have been started at this time point.
To fix this issue (and still keeping the functionality to monitor docker status with
systemd), we split the ExecStart process into an ExecStartPre part and an ExecStart
part. docker run is splitted into docker run -d then docker attach , while docker start
-a is splitted into docker start and then docker attach. In this way, we make sure
the downstream services are blocked until container is successfully started.
