Retry osquery instance launch until successful or shutdown requested #1952
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Outdated suggestions cannot be applied.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.
Closes #1937
When the osquery runner cannot launch an osquery instance, we currently return an error, which will shut down launcher entirely.
Looking over the logs and past issues we've investigated, I saw two primary errors: 1)
timeout waiting for osqueryd to create socket
, indicating the osquery process did not start up, and 2)could not create an extension client
where the socket file does not exist or the connection is refused.In both of these cases, restarting launcher is overkill, and even detrimental to solving the issue. In some cases, we can see these errors happen when the current osquery version is old and not compatible with the current database; restarting launcher in this case is actively harmful because it resets the autoupdate delay, preventing a newer osquery version from being downloaded.
Instead, this PR implements retrying osquery instance launch until successful (or shutdown is requested).
I originally also wanted to implement
If osquery instance launch fails, also consider triggering an autoupdate check for osquery
. However, I realized this part of the criteria will be significantly easier to implement once #1412 is complete, so I'm putting it off for now.