Handled recovery to start from lowest seqno in the translog based on retention lease #1217
Conversation
Signed-off-by: Sai Kumar <karanas@amazon.com>
|
✅ Gradle Wrapper Validation success 3a293c7d37e8d59e43d9381c8b3c6e1456713b70 |
|
❌ DCO Check Failed 3a293c7d37e8d59e43d9381c8b3c6e1456713b70 |
saikaranam-amazon
force-pushed
the
karanas-os-ccr-recovery
branch
from
3a293c7 to
6c546a3
Compare
|
✅ Gradle Wrapper Validation success 6c546a3 |
|
❌ DCO Check Failed 6c546a3 |
|
✅ Gradle Precommit success 3a293c7d37e8d59e43d9381c8b3c6e1456713b70 |
|
✅ Gradle Precommit success 6c546a3 |
| .mapToLong(RetentionLease::retainingSequenceNumber) | ||
| .min() | ||
| .orElse(Long.MAX_VALUE); | ||
| final long safeCommitSeqNo = Long.parseLong(safeCommitRef.getIndexCommit() |
There was a problem hiding this comment.
If I understand this right, tlog recovery is a mechanism to optimize serialization. Why would start sequence number change?
There was a problem hiding this comment.
We want to have the same set of tlog ops from the primary shard onto the replica shard during the peer recovery. This ensures the ops are available in tlog as well in newly recovered shard
Description
if pruning is enabled,
Issues Resolved
#1100
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.