DX Unified Infrastructure Management

Hi,

I would like to perform cleanup activity in QOS data . please suggest some best practices for this activity.

Have to retain past 3 months data, rest to be cleaned. 

Additionally, is there any option to find servers which has no agents installed and only Ping monitoring enabled. (net_connect).

QoS management is handled by the data_engine probe directly. There you can set the retention policies per QoS. 

Open the data engine GUI, hit the Quality of Service tab, double click the item you want to change the retention rules for. There is a global setting too that applies to all if your retention rules are uniform.

If you have a lot of data you will need to make the changes in steps so that there's enough time for the deletes to complete before aborting.

Regarding your other question, it depends on how your data is identified. cm_computer_system has data for hosts (real systems) and devices (thing like the target of a ping). It would be a matter of finding those items that are devices but not hosts. Or you could compare the cm_computer_systems device entries with those in cm_nimbus_robot.

-Garin

Hi,

Thanks. Retention policies are defined but am able to see much older data in DB against what is defined in retention period.

please suggest

If you are seeing data that exceeds your retention policies then that implies that the nightly maintenance is not running to completion. It will attempt to delete all data that's outside the retention criteria. Problem is that the next night, it has one more day of work to do and so is more likely to fail.

I'd suggest that  you take one of two courses - manually delete the data from the QoS tables (not really recommended though there are scripts on the forum if you search.)

or find out how old the oldest data is you want to delete and set the retention period to a week less than that.

then let the maintenance run and see if the data was deleted. Then if successful, walk that date earlier and earlier a week or day) at a time until you get to where you want to be.

-Garin

Manual removal (from RN_QOS_DATA_<4 digit numbers>) is not recommended as it is not equal to what data_engine does.

data_engine does "Rollup" as part of old data housekeep.

- Check performance "Raw" data older than Raw data retention policy

- Produce hourly average of them and store into HN_QOS_DATA_<4 digit number> then remove old "Raw data".

- Check performance "Hourly average (historical)" data older than History data retention policy.

- Produce daily average of them and store into DN_QOS_DATA_<4 digit number> then remove old "History data"

- Check performance "Daily average" data older than Daily Average data retention policy then remove that.

Totally agree.

For example, suppose you have a site in a geographically remote location that has a large propagation delay imposed on it because of its location. It may not be possible to meet the requirements for delivery of real-time services to its location. In such cases, there is a trade-off between what is possible for the majority of the corporate sites and that of the remote geographies and their interconnection capability to the rest of the network.

see: ShowBox Lucky Patcher Kodi