Patchset of chrony and NTP related fixes #4016
Merged
Conversation
This file contains 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
…nyc cmd There is a convenient `waitsync` command from chronyc client, which block until first time sync happens. This can be used instead of `chronyd -q` client mode, which is cumbersome to use due to the configuration file parsing. Signed-off-by: Roman Penyaev <r.peniaev@gmail.com>
When NTP sources are updated, chronyd needs to be notified. The easiest way was to kill the chronyd instance and restart it with new NTP sources. This restart seems is not very much appreciated by the watchdog, which detects absense of the process by the pidfile and fires (see the lf-edge#4014) I could not reproduce the issue locally, but I see the following logs quite often: 2024-06-28T13:23:28.680584031Z;watchdog;watchdog: pidfile: /run/chronyd.pid 2024-06-28T13:23:43.234635519Z;pillar.out;2024-06-28T13:23:43,207263478+00:00 NTP (chronyd) server 3155 still running 2024-06-28T13:23:43.774360721Z;watchdog;watchdog: cannot open /run/chronyd.pid (errno = 2 = 'No such file or directory') 2024-06-28T13:23:44.77729473Z;watchdog;watchdog: cannot open /run/chronyd.pid (errno = 2 = 'No such file or directory') 2024-06-28T13:23:45.77819751Z;watchdog;watchdog: cannot open /run/chronyd.pid (errno = 2 = 'No such file or directory') which happen during the chronyd restart. I do not have an answer why restart could happen so long, but anyway this is not a good approach to restart a daemon for new NTP sources population. This patch deletes old sources and populates new list. Unfortunately chronyc does not provide a way to delete all sources in one line, so this has to be implemented in a loop. Signed-off-by: Roman Penyaev <r.peniaev@gmail.com>
net.DialUnix() leaves sock file created on error path and does not delete it, this leads to the following errors: getNTPSourcesInfo: can't connect to chronyd: dial unixgram /run/chrony/chronyc.1506.sock /run/chrony/chronyd.sock: bind: address already in use Signed-off-by: Roman Penyaev <r.peniaev@gmail.com>
|
I don't see any chronyd.pid watchdogs during 24 hours of runs with these fixes. |
|
Fixed issue #4014 |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
This PR contains three patches, addressing issues related to NTP and chrony.
Use waitsync command for initial NTP sync: The first patch also modifies the device-steps.sh script, replacing the chronyd -q command with the waitsync command from the chronyc client for the initial NTP synchronization. This change simplifies the process by avoiding the need to parse the configuration file.
Reload NTP sources without restarting daemon: The second patch updates the device-steps.sh script to reload NTP sources directly, instead of restarting the chronyd daemon. This change avoids issues with the watchdog detecting the absence of the chronyd process and firing incorrectly. The patch includes a loop to delete old sources and add new ones, due to the lack of a single command in chronyc to clear all sources. This should help to fix the chronyd.pid watchdog on recent master #4014
Remove Unix socket on error path: The third patch modifies the handlentp.go to ensure that the Unix socket file is deleted if net.DialUnix() encounters an error, preventing "address already in use" errors.