diff options
| author | Martin Schwenke <martin@meltin.net> | 2011-01-14 09:40:11 +1100 |
|---|---|---|
| committer | Martin Schwenke <martin@meltin.net> | 2011-08-12 14:16:17 +1000 |
| commit | 1d71dd08e36414cdd4aac34ac21c247879447f28 (patch) | |
| tree | d561d050e7726d90e3196580f49a455941ea0866 | |
| parent | 398116ff295856dd5eccaa37acf4c7727501e8e8 (diff) | |
| download | samba-1d71dd08e36414cdd4aac34ac21c247879447f28.tar.gz samba-1d71dd08e36414cdd4aac34ac21c247879447f28.tar.xz samba-1d71dd08e36414cdd4aac34ac21c247879447f28.zip | |
Eventscripts: change failure counts and behaviour for statd and nfsd.
We reduce the number of failures before attempting a restart.
However, after 6 failures we mark the cluster unhealthy and no longer
try to restart. If the previous 2 attempts didn't work then there
isn't any use in bogging the system down with an attempted restart on
every monitor event.
Signed-off-by: Martin Schwenke <martin@meltin.net>
(This used to be ctdb commit f654739080b40b7ac1b7f998cacc689d3d4e3193)
| -rwxr-xr-x | ctdb/config/events.d/60.nfs | 9 |
1 files changed, 6 insertions, 3 deletions
diff --git a/ctdb/config/events.d/60.nfs b/ctdb/config/events.d/60.nfs index 43d5f665db..2b4c158d68 100755 --- a/ctdb/config/events.d/60.nfs +++ b/ctdb/config/events.d/60.nfs @@ -76,13 +76,16 @@ case "$1" in p="rpc.statd" which $p >/dev/null 2>/dev/null && \ nfs_check_rpc_service "statd" \ - -ge 10 "verbose restart" + -ge 6 "verbose unhealthy" \ + -eq 4 "verbose restart" \ + -eq 2 "restart:bs" # check that NFS responds to rpc requests if [ "$CTDB_NFS_SKIP_KNFSD_ALIVE_CHECK" != "yes" ] ; then nfs_check_rpc_service "knfsd" \ - -ge 15 "verbose restart unhealthy" \ - -eq 10 "restart:bs" + -ge 6 "verbose unhealthy" \ + -eq 4 "verbose restart" \ + -eq 2 "restart:bs" fi # check that lockd responds to rpc requests |