Details
-
Bug
-
Resolution: Fixed
-
Critical
-
Cheshire-Cat
-
7.0.0-5032
-
Untriaged
-
-
1
-
Unknown
Description
Steps:
1. Create a 1 kv, 1 query, 3 index node cluster
2. Create a bucket and 10 collections
3. Create 3000 indexes and wait for them to come online.
4. Drop Bucket
5. Rebalance out all indexer nodes without dropping indexes.
Rebalance exited with reason {service_rebalance_failed,index,
|
{agent_died,<22084.10695.13>,
|
{linked_process_died,<22084.10696.13>,
|
{no_connection,"index-service_api"}}}}.
|
Rebalance Operation Id = e8b1b8e009ec6fa297c40795ffd64438
|
Attachments
Issue Links
- backports to
-
MB-46523 [BP MB-45919 to 6.6.4] Indexer agent died on rebalance out all the indexer nodes while there are 3000 indexes and no data in KV and indexes
- Resolved
For Gerrit Dashboard: MB-45919 | ||||||
---|---|---|---|---|---|---|
# | Subject | Branch | Project | Status | CR | V |
152398,5 | MB-45919 Part 1: Add timing and logging of slow indexer msg processing | unstable | indexing | Status: MERGED | +2 | +1 |
152621,4 | MB-45919 Part 2: Add indexer high-priority message channel | unstable | indexing | Status: MERGED | +2 | +1 |
153036,6 | MB-45919 Part 3: Reduce O(N^2) work to O(N) for deleted indexes | unstable | indexing | Status: MERGED | +2 | +1 |