Check for NULL dbinfo, reset node seq#5960
Conversation
| hndl->connected_host = -1; | ||
| hndl->retry_all = 1; | ||
| newsql_disconnect(hndl, hndl->sb, __LINE__); | ||
| hndl->node_seq = 0; |
There was a problem hiding this comment.
set node seq to 0 since we did dbinfo response. bb does this
Signed-off-by: Salil Chandra <schandra107@bloomberg.net>
| parse_dbresponse(dbinfo_response, hndl->hosts, hndl->ports, &hndl->master, &hndl->num_hosts, | ||
| &hndl->num_hosts_sameroom, hndl->debug_trace, &hndl->s_sslmode); | ||
| cdb2__dbinforesponse__free_unpacked(dbinfo_response, NULL); | ||
| hndl->node_seq = 0; |
There was a problem hiding this comment.
I think we need to set node seq = 0 here in retry_queries too since it's dbinfo response? bb doesn't check for type dbinfo response in retry_queries, but it probably should I think
Signed-off-by: Salil Chandra <schandra107@bloomberg.net>
roborivers
left a comment
roborivers
left a comment
There was a problem hiding this comment.
Cbuild submission: Success ✓.
Regression testing: Success ✓.
The first 10 failing tests are:
sc_resume_logicalsc_generated **quarantined**
consumer_non_atomic_default_consumer_generated **quarantined**
longreq_stats
reco-ddlk-sql [timeout] **quarantined**
roborivers
left a comment
roborivers
left a comment
There was a problem hiding this comment.
Cbuild submission: Success ✓.
Regression testing: Success ✓.
The first 10 failing tests are:
sp_snapshot_generated
triggersc_latency
consumer_non_atomic_default_consumer_generated **quarantined**
reco-ddlk-sql [timeout] **quarantined**
2 participants
See commit by commit
The second commit is changes I've added that are not ports. But if we do a dbinfo response, I believe we need to reset these structures since they correspond to a stale host list