rust/neon
1
0
Fork 0
mirror of https://github.com/neondatabase/neon.git synced 2026-01-07 05:22:56 +00:00
Code Issues Packages Projects Releases Wiki Activity

Don't generate core dump when walproposer intentionally panics.

Browse Source 
Walproposer sometimes intentionally PANICs when its term is defeated as the
basebackup is likely spoiled by that time. We don't want core dumped in this
case.
This commit is contained in:
Arseny Sher
2023-11-27 15:40:01 +03:00
committed by Arseny Sher
parent 9bdc25f0af
commit e1935f42a1
3 changed files with 40 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

25
pgxn/neon/neon_utils.c
View File

@@ -1,3 +1,6 @@
#include <sys/resource.h>
#include "postgres.h"
#include "access/timeline.h"
@@ -114,3 +117,25 @@ pq_sendint64_le(StringInfo buf, uint64 i)
memcpy(buf->data + buf->len, &i, sizeof(uint64));
buf->len += sizeof(uint64);
}
/*
* Disables core dump for the current process.
*/
void
disable_core_dump()
{
struct rlimit rlim;
#ifdef WALPROPOSER_LIB /* skip in simulation mode */
return;
#endif
rlim.rlim_cur = 0;
rlim.rlim_max = 0;
if (setrlimit(RLIMIT_CORE, &rlim))
{
int save_errno = errno;
fprintf(stderr, "WARNING: disable cores setrlimit failed: %s", strerror(save_errno));
}
}

1
pgxn/neon/neon_utils.h
View File

@@ -8,5 +8,6 @@ uint32 pq_getmsgint32_le(StringInfo msg);
uint64 pq_getmsgint64_le(StringInfo msg);
void pq_sendint32_le(StringInfo buf, uint32 i);
void pq_sendint64_le(StringInfo buf, uint64 i);
extern void disable_core_dump();
#endif /* __NEON_UTILS_H__ */

15
pgxn/neon/walproposer.c
View File

@@ -35,6 +35,8 @@
*
*-------------------------------------------------------------------------
*/
#include <sys/resource.h>
#include "postgres.h"
#include "libpq/pqformat.h"
#include "neon.h"
@@ -1069,6 +1071,12 @@ DetermineEpochStartLsn(WalProposer *wp)
if (!((dth->n_entries >= 1) && (dth->entries[dth->n_entries - 1].term ==
walprop_shared->mineLastElectedTerm)))
{
/*
* Panic to restart PG as we need to retake basebackup.
* However, don't dump core as this is kinda expected
* scenario.
*/
disable_core_dump();
walprop_log(PANIC,
"collected propEpochStartLsn %X/%X, but basebackup LSN %X/%X",
LSN_FORMAT_ARGS(wp->propEpochStartLsn),
@@ -1445,7 +1453,12 @@ RecvAppendResponses(Safekeeper *sk)
if (sk->appendResponse.term > wp->propTerm)
{
/* Another compute with higher term is running. */
/*
* Another compute with higher term is running. Panic to restart
* PG as we likely need to retake basebackup. However, don't dump
* core as this is kinda expected scenario.
*/
disable_core_dump();
walprop_log(PANIC, "WAL acceptor %s:%s with term " INT64_FORMAT " rejected our request, our term " INT64_FORMAT "",
sk->host, sk->port,
sk->appendResponse.term, wp->propTerm);