Hi ! Got a trigger trouble.
I'm monitoring a backup server ( bacula ) :
Before each backup, i have this item updated by zabbix_trapper :
- "Kind of backup if running" ; possible value = 0 (nothing), F (Full backup in progress), I (Incremental...'gress ), D ( Differential )
After each backup, i have these items updated :
- Duration of last Full backup ( unit Min, could be in Sec )
- Duration of last Incremental backup ( unit Min, could be in Sec )
- Duration of last Differential backup ( unit Min, could be in Sec )
- Result code of last backup ( 0 or 1 for error )
- Kind of backup if running ( 0 which apply for "none in progress" )
I explain the trouble :-)
I already have a trigger on the backup_template :
{Backup_template:bacula.backup.result.last(0)}=1
It is the "Result code of last backup" : 1 = trigger on
So i'd like to know now if a backup is abnormally long.
I thought of this scheme :
"last value for Kind of backup if running" not equal "0" and "Kind of backup if running has received no data for more than 150% the average time it took for the last ten values received regarding the same kind of backup"
Ok let me say : It's a trigger in a trigger !
It could be 3 triggers ( one for abnormally long Full Backup is in progress, one for Incremental and the last for Differential )
Example for the "abnormally long Full backup is in progress" :
{Bacula_template:bacula.backup.status.str(F)}=1&{B acula_template:bacula.backup.status.nodata(({Backu p_template:backup.duration.full.avg(#10)})*1.5)}=1
But it simply don't work w/ err : [#10)})*1.5] is not a float or counter for function (avg)
Can someone help me to fix this please ? :-)
I'm monitoring a backup server ( bacula ) :
Before each backup, i have this item updated by zabbix_trapper :
- "Kind of backup if running" ; possible value = 0 (nothing), F (Full backup in progress), I (Incremental...'gress ), D ( Differential )
After each backup, i have these items updated :
- Duration of last Full backup ( unit Min, could be in Sec )
- Duration of last Incremental backup ( unit Min, could be in Sec )
- Duration of last Differential backup ( unit Min, could be in Sec )
- Result code of last backup ( 0 or 1 for error )
- Kind of backup if running ( 0 which apply for "none in progress" )
I explain the trouble :-)
I already have a trigger on the backup_template :
{Backup_template:bacula.backup.result.last(0)}=1
It is the "Result code of last backup" : 1 = trigger on
So i'd like to know now if a backup is abnormally long.
I thought of this scheme :
"last value for Kind of backup if running" not equal "0" and "Kind of backup if running has received no data for more than 150% the average time it took for the last ten values received regarding the same kind of backup"
Ok let me say : It's a trigger in a trigger !
It could be 3 triggers ( one for abnormally long Full Backup is in progress, one for Incremental and the last for Differential )
Example for the "abnormally long Full backup is in progress" :
{Bacula_template:bacula.backup.status.str(F)}=1&{B acula_template:bacula.backup.status.nodata(({Backu p_template:backup.duration.full.avg(#10)})*1.5)}=1
But it simply don't work w/ err : [#10)})*1.5] is not a float or counter for function (avg)
Can someone help me to fix this please ? :-)
Comment