OPA0: unit timed out

[VMSuser]

I have a VAX 4000-90 system. I notice the error count on the OPA0: keeps increasing whenever the device is allocated and then deallocated (when system manager has logged into the console and then logs out). I also notice that the device is timed out:

====

STPROA $ sh device/full opa0

Terminal OPA0:, device type LA36, is online, unit timed out, record-oriented
device, carriage control, error logging is enabled.

Error count 24 Operations completed 13619
Owner process "" Owner UIC [SYSTEM]
Owner process ID 00000000 Dev Prot S:RWLP,O:,G:,W:
Reference count 0 Default buffer size 132

STPROA $

======

I also notice that whenever I allocate OPA0: from another interactive session, and leave it allocated to myself (when no one is logged into the console), the error count stops increasing. Until I deallocate the device. At which point, the error count increases by 1. And then the error count keeps increasing until either I allocate the OPA0: device again, or someone else logs into the console, and stays logged into the console.

Has anyone seen this before? Is there a console patch available in VMS or console firmware available to fix this? The VAX 4000-90 is running VAX/VMS 5.5-2. I have another VAX 4000-90 running VAX/VMS 5.5-2H4, and the OPA0: device there does not have this problem. So I'm wondering if 5.5-2H4 is a possible fix.

Thanks in advance.

- Ron Beasley (**personal info erased**)

[Steven Schweda]

> [...] the error count on the OPA0: keeps increasing [...]

   Rate?  Who's doing the I/O which gets the error?  And you care
because...?

> Terminal OPA0:, device type LA36, [...]

   What is it _really_?

> Has anyone seen this before? [...]

   Not I, but I seldom use my hardware VAXes these days, and I'd be
unlikely to look at the error count on the console terminal when I do.
My VAX console tends to be an xterm on a Mac, with an ssh connection to
an IA64 system (rx2600, VMS V8.4-2L3), and a SET HOST /LAT connection
through a DECserver 90 to the physical console port on the VAX (4000
model 200, MicroVAX 3200?).  Might be different from your arrangement.

> [...] Is there a console patch available in VMS or console firmware
> available to fix this? [...]

   I know nothing, but I doubt it.  My guess at blame assignment would
be something you're doing, not a hardware/firmware problem.

> [...] the OPA0: device there does not have this problem. [...]

   Is the (non-OS) software identical?  Used identically?

> [...] So I'm wondering if 5.5-2H4 is a possible fix.

Again, I doubt it.  As I recall, -2H4 was a hardware-support
release (hence the "H"); I wouldn't bet that there were functional
changes.  (But it should be a harmless upgrade from plain -2, if you're
looking for an experiment to run.)