Imperva Cyber Community

communities_1.jpg
 View Only
  • 1.  What do you do when all DAS scans suddenly failed?

    Posted 06-17-2020 11:43
    This question has been asked more than one to our support team. I thought I would share it with you. @Ira Miga will answer it with her thoughts.  

    The following error was displayed, when we clicked on the "information" next to "Failed" status in the scan's history tab:

     

    The following error is displayed, only when running the scan (the scan almost instantly failed):

     




    #DataRiskAnalytics(formerlyCounterBreach)

    ------------------------------
    Christopher Detzel
    Community Manager
    Imperva
    ------------------------------


  • 2.  RE: What do you do when all DAS scans suddenly failed?

    Posted 06-17-2020 11:46
    @Christopher Detzel

    Here is the reason: 

    The main issue is the MX's results tablespace was almost full.

    The MX was constantly generating system events about the tablespace which is running low on space.

     

    • When tablespace running low a system event is generated - For example:The Tablespace RESULTS is running low on space. Warning Threshold of 85% has exceeded. Used space: 86%
    • When running a DAS scan, we check for tablespace and if it is below the threshold this system event is issued
    • This system event is aggregated, and when lots of scans running simultaneously there's a StaleObjectException on the SystemEventAggregationItem

     

    For some reason, these system events were sent at high rate (they're sent daily at 6:00 and on each scan run) this caused collisions between them (system events & results table space). This is the root cause of this issue.

    Solution / Workaround

    The bug was known to be resolved starting v12 - However, it should be noted that the fix evidently only includes an update to the Assessment Thresholds being pre-configured to 1,000.

    The steps below should be used as general troubleshooting steps when dealing with the errors mentioned above:

    1. Create a new policy that contains the same configuration as a policy that fails with the hibernate errors listed above.
      1. Next - run the scan
        1. If it fails - continue with steps 2 and onward.
        2. if it does not fail - Then chances are it has to do with what's currently stored for the scan 
          1. E.g. purging scans for the scan you replicated my allow it to continue without error.
    2. Change the assessment results purging definitions to be "by size" and keep 2-3 scan runs (Admin -> Maintenance -> Assessment Results Archive -> Purge Definitions)
    3. Reduce the result records threshold to 1000  (Admin -> System Definitions -> Management Server Settings -> Assessments)
    4. Consider deleting old unused scans



    ------------------------------
    Ira Miga
    Imperva
    Knowledge Engineer
    ------------------------------