yugabyte/yugabyte-db
Ver no GitHub[read-replicas] spurious logging (?): Expected replicas 3 but found 11 for tablet
Open
#3.235 aberto em 31 de dez. de 2019
area/docdbgood first issuekind/bugpriority/medium
Métricas do repositório
- Stars
- (8.229 stars)
- Métricas de merge de PR
- (Mesclagem média 17d 21h) (92 fundiu PRs em 30d)
Description
Jira Link: DB-1673 On a cluster with 3 primary replicas, and 8 read replicas, user reported the following:
The master's started logging a lot of Expected replicas 3 but found 11 for tablet halfway into the spike in errors.
The log line seems misleading (since 11 is indeed the expected number of replicas for the tablet in this setup).
Can we look to see if it is just spurious logging or a logic issue in master interpreting the read-replicas incorrectly?