E-Series Alerts and Monitoring

[israelmmi]

Do you know of any other methods by which we can monitor our E-Series? There is no SCOM Management Pack, not ActiveIQ integration, etc.

So when we have a system/disk/etc. failure, we need to solely rely on the e-mail alerts that the SANtricity OS sends, which gets many times swallowed up with all of the other million mails we get per day.

What do you do?

[elementx · Accepted solution]

There's Cloud Insights (the basic version is free, with acquisition VM on-prem and UI in the cloud), there's Grafana (open source, on-prem, download it from Github), there's Active IQ (cloud-based, mobile client available - for system risks https://docs.netapp.com/us-en/active-iq-1/reference_aiq_srd_health_summary.html#what-are-the-system-hardware-and-software-requirements).

[bretta]

Case automation and risks/health/wellness has been integrated into Active IQ Digital Advisor (formerly Active IQ). https://aiq.netapp.com There is a requirement to be sending back telemetry data, but that's available in all current version of SANtricity.

[PabloZorzoli]

hey, you can try out:

https://github.com/NetApp/eseries-perf-analyzer (supported by netapp)

or https://github.com/plz/E-Series-Graphite-Grafana

I also have a custom nagios check to report on broken hardware components, but haven't opensourced it If you really want to take a look at it, I can probably forward it to you 🙂

[israelmmi]

We are using Grafana but it doesn't send alerts.

AIQ is online based but doesn't send email alerts.

[elementx]

I don't have an E-Series in ActiveIQ to check, but I think ActiveIQ lets you configure email alerts in the Web UI.

Last time I checked (for SolidFire), it was possible, although not very friendly as one could enable email alerts for many different events and it wasn't clear what each of them meant.  Maybe it's simpler for E-Series, and essential alerts (link or controller down, dead disk and similar errors) shoudn't be many.

[NetApp_AU]

Hello,

For email alerts, you can follow this guide.

[israelmmi]

We know about email alerts.

We just receive hundreds of mails a day and it's quite easy (unfortunately) to miss a mail here and there. That's why we are looking for a real monitoring system integration so that the alert will never get closed or disappear until we take care of it.

[elementx]

> We are using Grafana but it doesn't send alerts.

https://grafana.com/docs/grafana/latest/alerting/notifications/

[israelmmi]

I was referring to visual alerts which stay put like in a central monitoring application. We don't need another email sent, since SANTricity already sends emails.

[KarelBP]

This is what I do:

1] Installed SANtricity Unified Manager (formerly Web Services Proxy)

2] Registered all E-Series storages there

3] Checking health a PowerShell with script every 15 minutes. The script does 2 REST API calls:

a] Lists all registered systems, their system IDs and health (Optimal/Needs Attention/Offline):

https://<unified_manager:port>/devmgr/v2/storage-systems

b] Gets failures without details:

https://<unified_manager:port>/devmgr/v2/storage-systems/<systemID>/failures?details=false

These failures are what you see in the GUI when it shows system is not optimal and Recovery guru tells you what to do. No details, only that system is not okay and what went wrong. This info is what goes into a ticketing system, because these are already actionable items, not like the events which first need to be interpreted to get some meaningful info. Of course you'll need to keep track of those which you already sent, you don't want to send the same thing every 15 minutes, that would be just as useless as getting your mail spammed with system events.

Hope it helps,

Karel