-1

I am trying to set up streaming replication in postgresql-10.0. i faced this issue.

Standbys are not supposed to receive any Transactional queries, if it does, postgres will throw out the following error and should not crash,

cannot execute DELETE in a read-only transaction

But im facing a crash in Standby when i issue a DELETE query. Attached the server logs,

postgres=# delete from time;
2018-08-09 12:51:12.965 IST [5166] ERROR:  cannot execute DELETE in a read-only transaction
2018-08-09 12:51:12.965 IST [5166] STATEMENT:  delete from time;
2018-08-09 12:51:12.965 IST [5166] ERROR:  cannot make new WAL entries during recovery
2018-08-09 12:51:12.965 IST [5166] WARNING:  AbortTransaction while in ABORT state
2018-08-09 12:51:12.965 IST [5166] ERROR:  cannot make new WAL entries during recovery
2018-08-09 12:51:12.965 IST [5166] WARNING:  AbortTransaction while in ABORT state
2018-08-09 12:51:12.965 IST [5166] ERROR:  cannot make new WAL entries during recovery
WARNING:  AbortTransaction while in ABORT state
2018-08-09 12:51:12.965 IST [5166] WARNING:  AbortTransaction while in ABORT state
WARNING:  AbortTransaction while in ABORT state
WARNING:  AbortTransaction while in ABORT state
2018-08-09 12:51:12.965 IST [5166] ERROR:  cannot make new WAL entries during recovery
2018-08-09 12:51:12.965 IST [5166] PANIC:  ERRORDATA_STACK_SIZE exceeded

ERROR:  cannot execute DELETE in a read-only transaction
ERROR:  cannot make new WAL entries during recovery
ERROR:  cannot make new WAL entries during recovery
ERROR:  cannot make new WAL entries during recovery
ERROR:  cannot make new WAL entries during recovery
PANIC:  ERRORDATA_STACK_SIZE exceeded
server closed the connection unexpectedly
    This probably means the server terminated abnormally
    before or while processing the request.
The connection to the server was lost. Attempting reset: 2018-08-09 12:53:39.180 IST [98163] LOG:  server process (PID 5166) was terminated by signal 6: Abort trap

I have no clue why standby is crashing. Does anyone have idea?

krithikaGopalakrishnan
  • 1,217
  • 10
  • 28

1 Answers1

0

Unless you have data corruption, that must be a PostgreSQL bug.

First, upgrade to 10.5 promptly.

If the bug persists, generate a stack trace and report the problem to the hackers mailing list or fill out the bug form.

Laurenz Albe
  • 209,280
  • 17
  • 206
  • 263
  • please can you have a look at this question https://stackoverflow.com/q/51758821/6633337 – Ratan Uday Kumar Aug 09 '18 at 11:18
  • @Laurenz Albe im noticing that, any query that results in a error statement, throws the same seg faults. for example, i try to query on a table thats dropped, I get an error message, "relation xxxx does not exist" and then i get this one. – krithikaGopalakrishnan Aug 13 '18 at 05:02