MicroStrategy - Todos


  1. See if the hipaaCheck can be moved to the phase after the wait phase. This seem not possible. MSTR suggest using setAsync and pollStatus, but I am not sure if this is any better than using rb.setMaxWait(-1). Can we get to the data using event? Can we get to the data using transform? Can we have multiple style / transform per report, and can we do this in the easily? The current transform that we are using for linking event ID is not flexible, but perhaps we can go back to exploring that option perhaps at the risk of having to deal with threading issue. See the "Extracting the report data" section from the "Ongoing investigation" page.
  2. It seems that the security filter check is not being performed on some dashboard reports. We need to confirm this, and make sure that it is run on all reports.
  3. The code that check for HIPAA violation seem to be slow on some reports. See if we can re-implement it as a transform instead of an add-on.
  4. Review the various governor / IServer settings to make sure that it is consistent and appropriate
  5. Fix the log statistics problem
  6. Automate purging of data from the statistics table and from the Enterprise Manager warehouse via a Command Manager script of via a Intelligence Server maintenance task
  7. Look at the cache directory on the Iserver, and open support ticket with MSTR to see how we can clean up stale cache.
  8. Configure administrative task to clean up history list messages and caches.
  9. Go through the installation page and do various tests mentioned there to see if the cluster is working appropriately, especially cache sharing and history list. Do this using the staging environment before doing it on production.
  10. Set up Enterprise Manager
  11. Do we need to set up Enterprise Manager for the DR environment? If we have a failover situation, do we need to have statistics for reports that were ran during that period?
  12. Security audit
    1. Should we log the SQL statement along with the name of the user?
    2. Should we also log the data (HTML, Excel, PDF)?
  13. Set up other MicroStrategy admin tools
    1. See if we can track the number of MSTR open web connections, MSTR database connections
  14. Set up IT360 dashboard
    1. Track database load
    2. Track IServer load
    3. Track IServer availability (restarts due to automatic update or manually)
  15. Regarding the issue where user not able to access MicroStrategy during the time that scheduled report is run:
    1. Perhaps we can re-configure the cluster such that one server is responsible for web traffic and the other server is responsible for scheduled jobs. This may not be ideal.
    2. Is there a way to restrict the number of concurrent scheduled jobs?
    3. Review job prioritization settings. Make sure that schedule jobs run with lower priority.
    4. Review the connection settings
    5. Review the admin manual
    6. Convert the 30 Classic SEM clients to QP
    7. Convert these freeform SQL reports into cube-based reports (including the admin reports)
    8. Performance benchmark to see how many scheduled reports that these servers can handle without making it not accessible. Performance benchmark for each setting that we change.
    9. Benchmark and limit the number of concurrent jobs associated with scheduled reports
    10. Rip out the scheduling interface from MicroStrategy, and implement our own interface and scheduling queue.
  16. Regarding not having enough environments for QA:
    1. We can possibly use different project names for each environment, I am afraid that this will increase overhead on the configuration side.
    2. We can add more environment to match with the application.
  17. Set up log collection and log viewer to help with debugging
  18. Monitor Tomcat availability. If I stop Tomcat, I should receive an alert.
    1. How can I monitor a particular web application / URL?
  19. Monitor the Intelligence servers. If I stop the Intelligence server, I should receive an alert.
  20. Create a dashboard showing the daily usage for all the nodes. It should include memory usage, CPU usage.
  21. Create a report that contains the daily usage for the previous day and email it to me.
  22. Configure SMTP.
Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-ShareAlike 3.0 License