[action] [PR:25506] [rsyslog] Fix Trixie restart delay with conditional restart#26576
Open
mssonicbld wants to merge 1 commit intosonic-net:202511from
Open
[action] [PR:25506] [rsyslog] Fix Trixie restart delay with conditional restart#26576mssonicbld wants to merge 1 commit intosonic-net:202511from
mssonicbld wants to merge 1 commit intosonic-net:202511from
Conversation
#### Why I did it On Debian 13 (Trixie), `rsyslog.service` includes systemd sandboxing directives (`PrivateTmp`, `ProtectSystem`, `ProtectKernelTunables`, etc.) that add ~4 seconds of overhead per restart due to namespace setup/teardown. This causes syslog-dependent tests to fail by missing log messages during the extended restart window. Fixes sonic-net#25382 #### How I did it Instead of always restarting rsyslog after generating the config, we now: 1. Generate the config to a unique temp file (`mktemp`) 2. Compare with the existing `/etc/rsyslog.conf` using `cmp -s` 3. **Config changed** (or first boot): install the new config and restart rsyslog 4. **Config unchanged**: send `SIGHUP` to rsyslog to re-open log files without a full restart The `SIGHUP` fallback is critical — it ensures rsyslog re-opens its file handles even when the config has not changed (needed after log rotation or `/var/log` remounts), while completely avoiding the 4-second namespace teardown/setup cycle. This preserves the upstream Trixie sandboxing directives. Additional hardening: - `mktemp` for unique temp files with `trap`-based cleanup on exit - Explicit first-boot handling (missing `/etc/rsyslog.conf`) - Error handling on `cp` failure (do not restart with stale config) #### How to verify it 1. Build a VS image with this change 2. Boot and verify rsyslog is running: `systemctl status rsyslog` 3. Run `config reload` — rsyslog should NOT restart (config unchanged), only SIGHUP: ``` journalctl -u rsyslog --since "1 min ago" # Should show rsyslog re-opening files, NOT a stop/start cycle ``` 4. Change a syslog config (e.g., add a remote server), run `config reload` — rsyslog should do a full restart 5. Run `syslog/test_logrotate.py` — both test cases should pass (SIGHUP re-opens files after /var/log remount) Signed-off-by: Sonic Build Admin <sonicbld@microsoft.com>
Collaborator
Author
|
Original PR: #25506 |
Contributor
|
/azp run |
|
Azure Pipelines successfully started running 1 pipeline(s). |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Why I did it
On Debian 13 (Trixie),
rsyslog.serviceincludes systemd sandboxing directives (PrivateTmp,ProtectSystem,ProtectKernelTunables, etc.) that add ~4 seconds of overhead per restart due to namespace setup/teardown. This causes syslog-dependent tests to fail by missing log messages during the extended restart window.Fixes #25382
How I did it
Instead of always restarting rsyslog after generating the config, we now:
mktemp)/etc/rsyslog.confusingcmp -sSIGHUPto rsyslog to re-open log files without a full restartThe
SIGHUPfallback is critical — it ensures rsyslog re-opens its file handles even when the config has not changed (needed after log rotation or/var/logremounts), while completely avoiding the 4-second namespace teardown/setup cycle. This preserves the upstream Trixie sandboxing directives.Additional hardening:
mktempfor unique temp files withtrap-based cleanup on exit/etc/rsyslog.conf)cpfailure (do not restart with stale config)How to verify it
systemctl status rsyslogconfig reload— rsyslog should NOT restart (config unchanged), only SIGHUP:config reload— rsyslog should do a full restartsyslog/test_logrotate.py— both test cases should pass (SIGHUP re-opens files after /var/log remount)Signed-off-by: Sonic Build Admin sonicbld@microsoft.com