Create and delete point in time for processing an index #2839
Conversation
graytaylor0
requested review from
chenqi0805,
engechas,
dinujoh,
kkondaka,
cmanning09,
asifsmohammed,
dlvenable and
oeyh
as code owners
|
Please comment out the failing tests, if they are failing because of sigv4 |
Signed-off-by: Taylor Gray <tylgry@amazon.com>
graytaylor0
force-pushed
the
PitCreateDelete
branch
from
b77da85
to
4aa79ca
Compare
A couple were failing from an unnecessary mock. I fixed those. No unit tests are failing from sigv4 issues. Manual testing was done to see the delete pit call fail with sigv4 |
|
|
||
| public class OpenSearchAccessor implements SearchAccessor, ClusterClientFactory { | ||
|
|
||
| private static final Logger LOG = LoggerFactory.getLogger(OpenSearchAccessor.class); | ||
|
|
||
| static final String PIT_RESOURCE_LIMIT_ERROR_TYPE = "rejected_execution_exception"; |
There was a problem hiding this comment.
This error type seems more generic than the PIT logic. Are there other cases where we can get a RejectedExecutionException unrelated to PIT?
There was a problem hiding this comment.
Probably but what is the alternative? Check the message contains something unique to the PIT message? Even for other exception we would log it and then give up the index and wait to start processing again, which is ok
…roject#2839) Signed-off-by: Taylor Gray <tylgry@amazon.com> Signed-off-by: Marcos_Gonzalez_Mayedo <alemayed@amazon.com>
Description
Create and delete point in time for an index to process it
enable_sigv4is set to true. I have not been able to debug, but it works properly whenenable_sigv4is false. I will create a separate issue to work on this bug in the future.Issues Resolved
related to #1985
Check List
By submitting this pull request, I confirm that my contribution is made under the terms of the Apache 2.0 license.
For more information on following Developer Certificate of Origin and signing off your commits, please check here.