How can I speed up bucket file processing for Spectrum Report Manager?
Article ID: 95528
Updated On:
Products
Issue/Introduction
Environment
Resolution
Here are some ways to increase the speed of bucket file processing.
1. From the OneClick console, Administration tab, Report Manager, Preferences, make sure that the Data Retention Policy is set to All Data and not Purge or Archive as when SRM is purging or archiving data, it is not inserting it into the Reporting database. After the bucket files have been caught up, this setting can be changed back.
2. If there are more then 1 landscape that is being reported on, you can temporarily disable the Event Processing on other landscapes so that the landscape that has fallen behind can catch up. You can disable "Event Processing" from the OneClick Administration page, Administration tab, Report Manager, SPECTRUM Status and uncheck the landscapes, so that only the behind landscape is being processed.
3. In Spectrum, we have an option to ignore the events from processing and these changes need to be done on a xml file.
Login to the SRM Server and go to $SPECROOT\tomcat\webapps\spectrum\WEB-INF\repmgr\config folder, make a backup copy and open the event-processing-filter.xml file.
Then we can add the event-type needs to be ignored. Please see the below example to do the same.
Example:
<ignore>
<event-type>0x10d66</event-type>
<event-type>0x10d67</event-type>
<model></model>
<model></model>
</ignore>
Please ignore the model field in this case.
Note: By doing this way all the port status events will be ignored on all the devices. Any future data collected by Report Manager will not contain these 0x10d66 and 0x10d67 events.
4. You can create an index on the SRM event table. Once logged into mysql and inside the reporting database you only need to run this command:
create index idx_model_key_time_type on event( model_key, type, time );
Depending upon the size of the current reporting database it may take some time to complete. Afterwards the time that the mysql database needs to do inserts into the table should greatly decrease which in turn will increase bucket file processing.
Feedback
