XIV crontab schedule for load_xivperf.sh


Using 7.2 version of stor2rrd.

There is a little confusion about the crontab entry for retrieving the XIV data.

When you run the connection test from the GUI and you still don't have the crontab entry, you get a warning message telling you to schedule load_xivperf.sh every 5 minutes.

But on the installation instructions, in the website IBM Monitoring: FlashSystem Storwize SVC XIV DS8K A9000 GPFS (stor2rrd.com), you are prompted to schedule it every 10 minutes -and definitely not less-.

Which scheduling should we use?

Thank you in advance,



  • Hi,

    use 10 minutes for XIV
  • Thank you @Pavel, we'll do that.

    Would you mind to elaborate? What's the reasoning behind?
  • XIV is not able to get all necessary data in 5 minutes.
    Well, it was the issue in the past, perhaps newer models are capacble to do it especially if there is no thousads of volumes.

    You can try even 5 minutes, you will see ...

  • Thx @Pavel

    So, you use 10min for not overlapping the crontab calls to load_xivperf.sh, right?
    I've tested it. In my environment (few hosts, few volumes) it ends in 5 seconds.
    Therefore, I let it run every 5min the crontab.
    [stor2rrd@stor2rrd ~]$ time /home/stor2rrd/stor2rrd/load_xivperf.sh
    Waiting for 96972
    real    0m5.383s
    user    0m2.909s
    sys     0m1.456s
    [stor2rrd@stor2rrd ~]$ _

    Little off-topic question. The XIV is missing in the HEALTH STATUS page. Is that right?
  • ok, then 5mins is fine.

    healt status is not available through SMI-S APi we use.
    or at least it was not available in the time we developed it.
  • Hi @Pavel,

    Maybe they implemented it: Indication profile - IBM Documentation

    AlertIndications are used to draw attention of subscribing client applications to the occurrence of an event. Lifecycle Indications are implemented in all XIV® CIM Agent releases. Beginning with XIV CIM Agent release 11.0, AlertIndications are implemented.
Sign In or Register to comment.