We found on the primary that EMS is taking up a huge amount of disk space. This could lead to an outage. How do we stop it?

Document ID : KB000006619
Last Modified Date : 14/02/2018
Show Technical Document Details
Issue:

We received and CDM disk alarm on the primary UIM server today.
When we researched it we found that the EMS probe directory had become huge. Over 40 GBS!
This seems to have started after we enabled alarm_routing_services and started having alarms flow through EMS.
What is causing this and how can we correct it.


The large folders were as follows:
Nimsoft/probes/service/ems/db/alarm_lifecycle_manager/YYYY-MM/DD
Nimsoft/probes/service/ems/db/alarm_manager/YYYY-MM/DD

examples:
/service/ems/db/alarm_manager/2017-04/23:
total 18G

./service/ems/db/alarm_manager/2017-04/24:
total 12G

./service/ems/db/alarm_lifecycle_manager/2017-04/24:
total 6.4G

./service/ems/db/alarm_lifecycle_manager/2017-04/23:
total 9.1G

Environment:
UIM 8.47EMS 8.45
Cause:

EMS stores all alarms in folders under the db directory.
This is similar to the NAS sql lite *.db files that hold the history alarms.
This is by design and expected.

Resolution:

You can reduce the amount of history the EMS probe stores but currently can not disable this all together.

set the following in the setup section for EMS Probe
<setup>
storage_retention_days = 1
enable_transaction_logging = false
</setup>

 

this will tell EMS to only store ONE day of alarm history.
Deactivate EMS and delete everything from the Nimsoft/probes/service/ems/db/ folder and then activate the probe.
This should cause the probe to only store alarms for the current day and one previous day.