Event Manager for VNX Block - Pool utilization

"Subscribed" is different then "Used". When you add data to the LUNs, you're filling up the LUNs and decreasing the available capacity of the Pool. As the LUNs get full with data, they will hit a limit for "Used" capacity and will send the alert. This is explained in more detail in the Help on the array. If you open the Pool to the Properties you will find the limits listed on one of the tabs. While you're on that tab, click on Help and it will explain what each item is.

glen

Thanks Glen. I need to mention that VP is not enabled. All LUNs are "thick" provisioned. The help file shows this:

"The consumed capacity of a pool that triggers an alert if the pool is oversubscribed. A pool may be oversubscribed if it contains thin LUNs or snapshots that subscribe to more capacity than is physically available in the pool. The default value is 70%. EMC recommends that you set the value somewhere between 50% and 75%. Regardless of the value that you set, the storage system automatically triggers an alert when the consumed capacity reaches the 85% full threshold value and, rather than trigger a new alert, updates the alert with each 1% increase in consumed capacity."

This implies if I create a thick LUN of 100 GB size it will "cunsume" a 100GB and so on. Now I have 97% consumed and the threshold is set to 50%. But no alert is being triggered.

Is my understanding correct?

It's my understanding that you should have seen an alert when you got over 85%, but with only Thick LUNs in the Pool, there is no way that the Pool can grow beyond what you originally configured it for and that may be why you didn't see an alert. If there were Thin LUNs or VNX Snapshots, then it would be possible to exceed the physical capacity if the Thin LUNs were oversubscribed or if you have VNX Snaps and that would trigger the alert. It seems like it's an "if" test -

if Thin oversubscribed and/or VNX Snaps, then alert

The real purpose is to warn you that you're getting close to a full condition. You really do not want to get the Pool completely full. See page 4 in the attached.

glen

Thanks again Glen. That's what we want to get. An alarm if it comes to a close to full condition. The alarm should be displayed at the Management UI and it should be forwarded to our internal ticket system via SNMP.

Regards

Dieter

I tested in my lab and found that the only time I see an alert is if I have a Thin LUN that is oversubscribed and I add files to the LUN that pushes it over the 70% full mark,.

You can setup an Notification - the general code for Virtual Pools is 712d xxxx - those will be the Event Codes for Pool events including the 70% alert.

glen

Glen,

thank you very much for the effort. I'll take this in consideration and update my template.

Is there a document which describes the alerts for 5.32? Searching within Event Manager is not possible and therefore somehow strange to find the right one.

Dieter

See the attached - the Unisphere Release notes have the events listed near the end.

glen