Device FGB0: registering errors

[Jasper_15]

Greetings,
I did a show error, and noticed that my FGB0: device is racking up errors. I am not sure if this device being used and I could not able to search my VMS for this particular device. I haven't notice this device for sometime. I've displayed the device in full for everyone to see. Please direct me where and how to dectect this device. Also, how would you go about redirecting this device to use a different path. Thanks!!


Device FGB0:, device type KGPSA Fibre Channel, is online, shareable, error logging is enabled.

Error count 2 Operations completed 0 Owner process "" Owner UIC [SYSTEM]
Owner process ID 00000000 Dev Prot S:RWPL,O:RWPL,G,W
Reference count 0 Default buffer size 0
Current preferred CPU Id 2 Fastpath 1
Current Interrupt CPU Id 2
FC Port Name 1000-0000-C936-6E92 FC Node Name 2000-0000-C936-6E92

[Thomas Ritter · Accepted solution]

Jasper, provide an overview of your system: vms version, host types, interconnect topology and storage solution.

Are the errors affecting a service ?

If applicable, what is the output from $MC SCACP show channels ?

[Jasper_15]

I am sorry. This is not part of a cluster environment. This is stand-alone server with SAN storage running OpenVMS 7.3-2

Thanks,

[Uwe Zessin]

It's a Fibre Channel adapter. FGB0 usually maps to the console device PGB0, but you should compare the WWNs to be sure.

[Wim Van den Wyngaert]

Try
$ mc lancp show dev fgb/count
and check for abnormal things.
It could be that errors are logged by the cluster protocol when a simple "buffer unavailble" is found.

BTW : on my system I have 48 receive data length errors but no errors logged. But the other cluster meber has no errors.

Wim

[Rinkens]

Hi Jasper

You can do several things to check your system.

$ show dev pga0
$ show dev pgb0

and see if the errors are rising.

Next waht you can check is the

$ show dev/multi d

Here you can see howmany paths there are to
you san disks.

$1$DGA24: ShadowSetMember 0 9/ 9 PGB0.5000-1FE1-0010-9722
$1$DGA25: ShadowSetMember 2 9/ 9 PGB0.5000-1FE1-0010-98D2
$1$DGA26: ShadowSetMember 0 9/ 9 PGA0.5000-1FE1-0010-9722

The 9/9 shows 9 paths and 9 paths available.

Now you take one of your disks and do a full

$ show dev/fu $1$dg100 (exmaple)

You know can see which path is current.

If there is no path anymore on the pgb0. Then maybe it is broken.

Look also in the errlog.sys file with dia if possible, or check the other site off the pgbo cable (your switch or hsg80) if it is reporting faults.

Regards KOr

[Jorge Cocomess]

QBN-RAVEN >show dev/multi d
Device Device Error Current
Name Status Count Paths path
$1$DGA32: (CARNARY) Mounted 1 4/ 4 PGA0.5000-1FE1-000B-B1C4
$1$DGA52: (CARNARY) Mounted 0 4/ 4 PGB0.5000-1FE1-0006-2831
$1$DGA53: (CARNARY) Mounted 1 4/ 4 PGA0.5000-1FE1-0006-2834
$1$DGA54: (CARNARY) Mounted 0 4/ 4 PGB0.5000-1FE1-000B-B1C3
$1$DGA55: (CARNARY) Mounted 0 4/ 4 PGB0.5000-1FE1-000B-B1C1
$1$DGA56: (CARNARY) Mounted 1 4/ 4 PGA0.5000-1FE1-000B-B1C2
$1$DGA73: (CARNARY) Mounted 0 4/ 4 PGB0.5000-1FE1-000B-B1C3
$1$DGA74: (CARNARY) Mounted 0 4/ 4 PGA0.5000-1FE1-000B-B1C2
$1$DGA75: (CARNARY) Mounted 0 4/ 4 PGA0.5000-1FE1-000B-B1C2
$1$DGA76: (CARNARY) Mounted 1 4/ 4 PGA0.5000-1FE1-000B-B1C4
$1$DGA77: (CARNARY) Mounted 0 4/ 4 PGB0.5000-1FE1-000B-B1C1
$1$DGA199: (CARNARY) Online 2 2/ 2 PGB0.5000-1FE1-000B-B1C1

Seem like everything is working fine and I am seeing all 4 paths.

I logged on to the SAN switch, I am not seeing any errors. Where do I go from here?

Help!!

Thanks!!

[Rinkens]

Hi

If you have a service contract you can make a call at HP and send the errorlog.sys to HP, they can analyse the errorlog.sys for you.

Maybe you have to install some patches or, upgrade your firmware.

$prod show hist

and

$type sys$update:vmsinstal.history.

I would go for the first option.

Regards Kor

[Jorge Cocomess]

Hi,

How do I get into the errlog.sys to look for errors on my FGB0: device?

Thanks,
J

[Rinkens]

$ Set def sys$errorlog
$ Dir/dat=(cre,mod) errlog.sys

Look if it is not very new version (one with a date where a error occured). Otherwise it can be that you have to use the errlog.old (sometimes scripts rename the errlog.sys to errlog.old, otherwise the errlog.sys can be to big).

You can not type this file because it is a binary.

Sometimes you use one of the following below.
Type off hardware version is leading in this.

1)
$ ana/error errlog.sys/out=w.w
2)
$ set def sys$errorlog
$ dia/out=w.w/sine=1-dec=2005
$
3)or webes

Greetings Kor