Tape robot GG devices disappear on one node but not other

[Clark Powell]

Two identical (hardware & software) nodes, OVMS 8.3, in a cluster, boot after a full power off and on one node the MSL6060a juke box controller device does not appear:
ALPHAD> sho dev gg
Device Device Error
Name Status Count
$1$GGA5: Online 0
$1$GGA6: Online 0
while on the other they do appear:
ALPHAC> sho dev gg
Device Device Error
Name Status Count
$1$GGA5: Online 0
$1$GGA6: Online 0
$2$GGA0: Online 0
$2$GGA1: Online 0
Although the two nodes have separate system disks, the SYS$SYSTEM:SYS$DEVICES.DAT file is the same on both (except for time stamp of course.)
Running autoconfig produces the following result:
ALPHAD> SYSMAN IO AUTO/LOG
%SYSMAN-I-TIMEVAL, timeout value is 0 00:00:15.00

%SYSMAN-I-OUTPUT, command execution on node ALPHAD
%IOGEN-I-PREFIX, searching for ICBM with prefix SYS$
%IOGEN-I-PREFIX, searching for ICBM with prefix DECW$
%IOGEN-I-SCSIPOLL, scanning for devices through SCSI port PKA0
%IOGEN-E-LOADERR, error configuring device SRA0, driver SYS$SRDRIVER.EXE
-RMS-E-FNF, file not found
%IOGEN-I-FIBREPOLL, scanning for devices through FIBRE port PGA0
%IOGEN-I-FIBREPOLL, scanning for devices through FIBRE port PGB0

Is the missing SR driver the culprit? Or is it something else?

No changes have been made since the previous boot. Fibre HBA's were upgraded on node with missing GG device in May but many boots since then have occurred. Disk are from EVA 5000 so fabric is functional and MGA devices are there as normal.

Robot (GG device) was functional before full, planned power off. Before only one node at a time would have been shut down.

thanks
Clark Powell

[Duncan Morris · Accepted solution]

Clark,

SRDRIVER is unlikely to have anything to do with this. See Jur's response in this thread

http://forums11.itrc.hp.com/service/forums/questionanswer.do?threadId=1125469

[Clark Powell]

HP Storage support fixed this one. Apparently, when we changed the Fibrechannel HBA's last May the NSR adapted to the new HBA's WWID's. Then when during the last outage during which both the NSR and the two nodes of the cluster were powered down, the NSR was no longer able to communicate with new HBA's. We found that the old HBA's WWID's were defined in the NSR mapping. We deleted them. The NSR had also identified the new WWID's but would not use them until we mapped them. I have to say that I don't understand why it was working and then after the power shutdown it wasn't. I guess, just because something works, doesn't mean that it is set up correctly. But the good news is that I understand why it's working now...
Clark Powell

[Christopher Borg_1]

Hi Clark,

Are you the Clark who worked at USF? if yes please reply I have a message for you from a mutual friend.

BTW USF dissolved the Operations department & no longer uses VAX/Alpha

Thanks
cb

[Clark Powell]

Cris,
Yes, I'm the same. It was a good job with some very good co-workers. The equipment was so bad that it was really a great learning experience for me. We had a disk crash every month! I became like the surgeon who served in a MASH unit who knows from endless cutting under extreme circumstances, what's possible and what's not. That was worth the 60 or 70 hours a week I put in but I never let anybody work me that hard again, (unless they were paying by the hour, of course!) I hope you get this as I couldn't find an email in your profile. My email is in my profile. Click on my name and you can send an email.

Clark Powell

[Christopher Borg_1]

Clark,

I couldn't find the email in your profile, anyway Frank Dosh is trying to reconnect with you Frank is living in Cambodia. balifranki@gmail.com

Thanks
cb

[Clark Powell]

case close thanks