IT Services calculation

[hannibal20]

Hi,

I have a problem with SLA propagation. Steps taken:

1. created a custom key (based on file checksum)
2. created a simple trigger (fires off when file checksum changes, disaster level)
3. created a service linked with my trigger, no algorithm, not showing SLA
4. created a higher level service, show SLA, MIN algorithm
5. linked pt. 3. as a child to pt. 4

Now when I fire off my trigger, service form pt. 3 changes ststus to disaster, the status propagates to service from pt. 4, but time still is counted for the 'OK' time, not for the 'Problem' time.

Server log says:

15183:20070605:104210 Query failed: [insert into service_alarms(serviceid,clock,value) values(10,1181032930,5)] Duplicate entry '0' for key 1 [1062]

I'm using the 1.4

[hannibal20]

patch

this seems to solve the problem

*** src/libs/zbxdbhigh/db.c_orig        Tue Jun  5 12:52:57 2007
--- src/libs/zbxdbhigh/db.c     Tue Jun  5 13:03:58 2007
***************
*** 429,435 ****
                return SUCCEED;
        }
  
!       DBexecute("insert into service_alarms(serviceid,clock,value) values(" ZBX_FS_UI64 ",%d,%d)",
                serviceid,
                clock,
                status);
--- 429,436 ----
                return SUCCEED;
        }
  
!       DBexecute("insert into service_alarms(servicealarmid,serviceid,clock,value) values(" ZBX_FS_UI64 "," ZBX_FS_UI64 ",%d,%d)",
!               DBget_maxid("service_alarms","servicealarmid"),
                serviceid,
                clock,
                status);