How to kill a CMDB Health Dashboard job?

[rrataj]

Hi,

I have executed the CMDB Health Jobs, and unfortunately have seen no data on the Dashboard so I started to digg

to see information from CMDB Health I have enabled logging

I have disabled the CMDB Health jobs and on the next day hoped to execute them manually,

however looking into the logs I see:

Level	Message	Source	Created by
Information	CMDBHealth : RECOMMENDED time out. Exiting this processor run	CMDBHealth	system
Information	CMDBHealth : RECOMMENDED : Starting Processor with total time 1200	CMDBHealth	system
Information	CMDBHealth : RECOMMENDED time out. Exiting this processor run	CMDBHealth	system
Information	CMDBHealth : RECOMMENDED : Starting Processor with total time 900	CMDBHealth	system
Information	CMDBHealth : RECOMMENDED time out. Exiting this processor run	CMDBHealth	system

now when I am executing the jobs manually I do get such information in the log

Level	Message	Source	Created by
Information	CMDBHealth : CORRECTNESS: The metric is currently being processed, skipping this run	CMDBHealth	rafal.rataj.global
Information	CMDBHealth : COMPLIANCE: The metric is currently being processed, skipping this run	CMDBHealth	rafal.rataj.global
Information	CMDBHealth : COMPLETENESS: The metric is currently being processed, skipping this run	CMDBHealth	rafal.rataj.global

only the Relationship Score Calculation Job does not create any log information, which seems also weird, even if executed twice there is no information about the job

So my questions are:

1) is how to check if the CMDB Health Job is running ?

2) how to terminate the running job ?

Please note this is not a standard PA Job.

[rrataj]

In the Active Cluster Transactions I do see some strange job

URL: ASYNC: Script Job,

type: Scheduler,

Foreground: false

State:database

SQL statements:98,629

SQL statement

SELECT cmdb_health_processor_status0.`sys_id`

FROM cmdb_health_processor_status cmdb_health_processor_status0

WHERE cmdb_health_processor_status0.`metric` = 'c7d72b5137130200f212cc028e41f1bd'

AND cmdb_health_processor_status0.`status` = 'DRAFT'

ORDER BY cmdb_health_processor_status0.`batch_name` limit 0

,1 /* atosglobaldev004, gs:glide.scheduler.worker.1, tx:5f531f5d37c107009d3b861754990eda */

it looks like its called constantly each several minutes just from somewhere ?

additionally the table cmdb_health_processor_status is constantly updated, so something is running there ...

[manishm]

You must have a very big CMDB. We have approx 1 million CIs and processing of just one metric took upwards of 2 hours. I usually leave the job overnight and by the morning, we have some scores. Try enabling the parameters one by one (I have just duplicate check on), then we plan to do staleness, followed by correctness).

The fact that you got the "metric is being processed" tells you that the job is running.

See if this helps to terminate the transaction:

Viewing and Killing Active Transactions - ServiceNow Wiki

Manish

[rrataj]

Hi Manish,

thanks for the reply,

the CMDB is below 1 mln, hoewver the domain separation is a problem, we have a lot of domains.

Meantime I figured out how to kill the jobs.

One way is to set the processing statuses to 'Complete' however this this is just skipping some processing, if your lucky the job will stop.

You can see the running jobs in the schedules (all CMDB Jobs scripts contain   SNC.MetricProcessorScript[...] )

https://INSTANCE.service-now.com/sys_trigger_list.do?sysparm_query=state%3D1%5EscriptLIKESNC.MetricP...

the name is ASYNC: Script Job

go to Active Cluster Transactions

https://INSTANCE.service-now.com/v_cluster_transaction_list.do?sysparm_query=url%3DASYNC:%20Script%2...

and here you can kill the job