HPE EVA Storage

Loop B Operational State: Failed. clicking Enable does nothing

 
Matt_Hale
Occasional Advisor

Loop B Operational State: Failed. clicking Enable does nothing

We have inherited an old EVA. I think it's a 4000?  It has two HSV200 controllers, 8 disk enclosures, and no switches.  It has an annoying habit of randomly dropping fibre links on the disk enclosures, but it is cabled for full redundancy and so far this hasn't been more than an annoyance.  Reseating the cable always brings the link back up again.

This time however, two ports on the B loop failed in short order.  We have reseated both and the links have come back up as usual, however the Controller Properties screen in Command View still shows the Loop B Operational State as Failed.  Clicking the Enable button next to it produces an Operation Succeeded message but nothing changes, it remains failed.  Note that all the disks report Loop B State as Good - it only shows as failed on the controllers.

I am worried that if we experience another random link loss on the A loop in this state, we are going to lose access and possibly data.  Can anyone suggest how to to get Loop B back again?

Thanks very much.

 LoopB.png

3 REPLIES 3
Matt_Hale
Occasional Advisor

Re: Loop B Operational State: Failed. clicking Enable does nothing

Overnight the state of the B loop has changed to Good on the A controller.  It remains Failed on the B controller.  I really have no idea what is going on.

Matt_Hale
Occasional Advisor

Re: Loop B Operational State: Failed. clicking Enable does nothing

And now 48 hours after the original incident, both controllers are reporting the Operational State of Loop B is Good, with apparently no intervention.  Can anyone suggest why the state might change on it's own like that?  Especially after manually enabling the loop fails?  Or why it took so long?

veeyarvi
HPE Pro

Re: Loop B Operational State: Failed. clicking Enable does nothing

Hello Matt_Hale

This EVA 4000 (HSV200 is EVA4000) has mis behaving backend device and that is why we are seeing this behaviour.  It could be a drive, cable, IO module or even any of the controllers. 

When the device is misbehaving, the loop will be down (you will see the port errors for the respective controllers) and the loop cannot be enabled until the issue is fixed. And, since the loop is coming back automatically (the controllers will retry enabling the loop in the background periodically), it appears the issue getting cleared intermittently.

Only the analysis of the controller event logs can clearly tell what is happening. The command view EVA is of little help in troubleshooting intermittent loop issues.

You should reach support if the array is too critical for you since such issues may end up in a complete array down scenario if not treated in time. 


I am an HPE Employee
Accept or Kudo