Make sure other tasks are processed with tasklists that contain both bad xfrs (that cause reload to exit) and other tasks #390
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
This test is to make sure that the effects of other tasks are effectuated when they are processed together with other tasks in a single tasklist. The test does a reload for a single bad update (from xfr) together with an addzone in a single tasklist. The addzone should be effectuated.
So that the backup main has all other tasks processed when one of the transfers fails (causing an exit)
Directly without main or backup-main proxying for it.
... when configured to do so
wcawijngaards
approved these changes
Oct 22, 2024
tpkg/bad_xfr_with_other_tasks.tpkg/bad_xfr_with_other_tasks.test
Outdated
Show resolved
Hide resolved
|
Just for the record. A write to a pipe is guaranteed to be atomic when the written size is less then An alternative (simpler?) solution could be to use a single pipe to the I think that with such a solution the necessity for a "NOTIFY per zone" buffering mechanism becomes a bit more urgent though, since the serve children loose their individual pipe buffer. |
Grazie! Co-authored-by: Wouter Wijngaards <[email protected]>
k0ekk0ek
approved these changes
Oct 23, 2024
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Outdated suggestions cannot be applied.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.
The issue is that updating the database from transfers and other tasks that change the database (such as adding zones or changing patterns) are processed in so called task list. Especially applying transfers are executed batch-wise periodically as configured by xfrd-reload-timeout: (default every second). If a zone is added it is appended to the tasklist. If there would have been an "apply transfer task" before that other (adding a zone) task, then that will be executed first. If the "apply transfer task" fails, the reload will be aborted (because the database could be corrupt) and the effect of any tasks before or after the transfer are undone. This can lead to a situation that the transfer daemon (xfrd) thinks a zone is configured, but the zone is not configured in the server process (main).
This PR replaces PR #332 . It removes the additional fork by letting serve childs send NOTIFY messages directly to the xfrd process instead of letting main or backup-main proxy for it.