StarWind VSAN Free (KVM) – No Email Notification on RAID5 Degraded

[huangdog]

Installed version:

1.6.580.7361

I am using StarWind VSAN Free in KVM. While testing its software RAID and email notifications, I noticed that when RAID5 becomes degraded, no email notification is sent. The system only sends an email after the RAID fails completely and files become inaccessible.

I have already configured Event Notifications → Send E-mail via the StarWind Management Console on Windows for my StarWind VSAN Free in KVM setup. Despite this, there is still no notification when the RAID is degraded.

Is this expected behavior? Is there any way to receive email alerts when the RAID is degraded, before a complete failure occurs?

Thanks in advance for any advice!

I plan to use StarWind VSAN Free (KVM) as the hard disk and file management system for my home DIY NAS. I installed PVE on my DIY NAS, and then installed StarWind VSAN Free (KVM) in a VM on PVE.

[yaroslav (staff)]

Welcome to StarWind Forum!
StarWind CVM does not feature our custom email notification for MDADM. What you see in CVM is the general MDADM settings.
We plan to introuce MDADM monitoring later, in future builds.
For now, you can set up SNMP or monitor your system with Zabbix.

Event Notifications → Send E-mail via the StarWind Management Console

This bit sends only StarWind VSAN events. You can see the list here https://www.starwindsoftware.com/help/E ... dVSAN.html

[huangdog]

What is the root password in ssh?

Withou root permission , I can not config the snmp to monitor the raid status.

[yaroslav (staff)]

I can't disclose this information.
Again, you are welcome to use the credentials you've set during installation.
You can use sudo.

[huangdog]

Could future updates integrate email sending directly into StarWind VSAN? I don’t want to install a console on Windows just to send emails—it’s too cumbersome.

[yaroslav (staff)]

I hope there will be another straightforward means for monitoring. MADM, technically, is not the service part, so it might not be in the Management Console.
As a side note, you don't need to run Windows full-time for monitoring; just for setting.

[huangdog]

Storage is the core of StarWind. If there is no monitoring of Madam, causing the rescue opportunity for the soft RAID to be missed and leading to the entire soft RAID failing, then no matter how well StarWind is built, it will all be meaningless.

[yaroslav (staff)]

Good point. We are working on it. Yet, I don't agree with your take entirely.
MDADM and disks are non-StarWind products. Users are welcome to implement their flavor of third-party monitoring. SNMP or Zabbix are the first solutions that come to my mind; both are third-party solutions.
VSAN logging in its current state is not as bad as it might seem from this thread. True, the logging in its existing form does not indicate a disk or MDADM failure, yet VSAN still will log being unable to write to an *.img, which can be interpreted as an issue with StarWind's underlying storage, and send you an alert. On the other hand, such alerting will be excessive as it does not distinguish between actual disk failure and, say, disk halt due during the VM snapshots (we don't recommend doing snapshots of StarWind VMs or backing them up; see https://www.starwindsoftware.com/best-p ... practices/).
Speaking of precise disk monitoring, Disk monitoring is available for systems with ProActive Support (i.e., vHCA, HCA, and ProActive support).
Still, we are working on convenient MDADM monitoring for all users since the product is coming as an appliance.

[emelialucass]

If Madam is not monitored, resulting in a missed chance to rescue the soft RAID and ultimately causing its complete failure, then regardless of how well StarWind is designed, it will all be in vain.

[yaroslav (staff)]

Yes, it is not monitored out of the box. But you can set up monitoring there yourself with PRTG/Zabbix/etc., and you can set up alerting for MDADM specifically. So, the claim above is wrong: you CAN.
Degraded is not failed. But Degraded can become failed unless action is taken or the disk swap is mishandled.
If MDADM fails in a highly available shared storage configuration, you still have StarWind VSAN on another side(s).

[wiliamcorlin]

Hi, I’m new here but running a similar lab setup with StarWind VSAN on KVM.

For the MDADM monitoring part — since root access is restricted, has anyone tried setting up mdadm --monitor with email alerts via a non-root sudo config (like allowing specific mdadm commands without full root)? Or is SNMP/Zabbix really the only practical way in the current builds?
Just trying to find the lightest workaround without adding a full monitoring stack

[yaroslav (staff)]

You can do it with sudo.