Mount Verify

[rison so_1]

Since last Friday, one of my ES45 in cluster running OpenVMS 7.3-2 keeps doing "Mount Verify" on a few disk in the SAN. The other system in the cluster is OK. While the problem system is doing the "Mount Verify" every stops then comes back when it finished mounting. Any idea ?

[Wim Van den Wyngaert · Accepted solution]

Rwead the wise words of Volker :

a disk enter mount-verification, if an IO to that disk does not finish successfully. Some IO errors will be logged in ERRLOG.SYS, other IOs would be just retried and temporary error conditions would not be easily detectable.

The IO error can happen anywhere in the path between the disk driver and the actual disk. Some components in this path provide error logging (did you look at the HSG80 console ?), others don't.

Especially in a SAN, any problems in the fabric can cause temporary IO errors. Look at the description of the MVSUPMSG_INTVL SYSGEN parameter.

Do you see path switched messages in OPCOM ?

May these MV messages be related with any unusual IO activity on the system ? At specific times ?

Wim

[Wim Van den Wyngaert]

Also read my question (that is not solved but helps you understand MV).

https://forums1.itrc.hp.com/service/forums/questionanswer.do?threadId=1125853

Wim

[The Brit]

Can you confirm the type of storage (EVA/HSG)?
The implication from your post is that it is HSG80 storage ("... on a few disk in the SAN"). If this is true, then you could possibly have

1. a controller problem causing path switches (check whether the devices are pointed to a particular controller). Look at the Prefer = THIS or Prefer = OTHER status (not sure about the exact syntax, its been a while)

2. a shelf problem (check whether the devices are on a particular shelf).

If this storage is EVA, and units are presented from different EVA's, I have seen this situation occur when both controllers decide to reboot at the same time, i.e. if the second controller decides to reboot before the first controller has finished rebooting (although this normally resulted in the affected units eventually being dropped).

Dave.

[rison so_1]

Problem solved. Faulty EMU on a disk self.

Thank you all for your help.