XtremIO X2: Software module failure(s) and potential service disruption due to extensive logging on XtremIO clusters running software versions earlier than 6.2.0-81[2]


   Article Number:     531449                                   Article Version: 7     Article Type:    Break Fix 




XtremIO Family,XtremIO X2,XtremIO HW X2-R,XtremIO HW X2-R Encrypt Disable,XtremIO HW X2-S,XtremIO HW X2-S Encrypt Disable,XtremIO HW X2-T,XtremIO HW X2-T Encrypt Disable





An XtremIO X2 cluster experiences one or multiple software module failures, which may result in:   

  •         Impact to ongoing native replication activity on the cluster (due to failure of the software modules responsible for servicing replication traffic)     
  •         Service disruption on the host side (due to non-compliance with the relevant best practices contained XtremIO Storage Array Host Configuration Guide)     
  •         Service disruption due to cluster emergency shutdown (in the event of multiple software module failures, following which the cluster will either auto-recover or will require manual intervention in order to bring the cluster back up)     






An XtremIO software issue which may trigger one or multiple software module failures due to extensive logging of buffer overruns within the log files of one or more Storage Controllers (SCs) following the reception of massive I/O bursts by one or more of the target ports residing on the SCs in question.                                                           






A permanent fix involving less aggressive logging of buffer overruns is available within XtremIO software versions 6.2.0-81 and above.    
    As a general recommendation, host I/O requests should be equally distributed across all available target ports on XtremIO clusters, thereby preventing clusters from running out of resources for particular target ports and SC software modules by providing more channels and buffers to handle incoming I/O. If this is not the case (as can be confirmed by executing the show-targets-performance XMCLI command on the cluster in question), the necessary zoning changes should be made in order to accommodate the aforementioned recommendation.






The referenced software issue is tracked under XIO-59454 and is described within the latest XtremIO Storage Array XIOS Release Notes as follows:   
    An internal software module may time-out due to extensive message logging.