[Info-vax] UCX Portmapper fails with %RMS-E-RNF, record not found

[Hans Bachner]

Stephen Hoffman schrieb am 03.05.2018 um 20:05:
> On 2018-05-03 16:59:48 +0000, Hans Bachner said:
>
>> you are my last resort :-)
>
> So you have no backups,
we *do* have backups, I mentioned one in the base note, and older ones 
are around as well.

What I tried to find out is *what* needs to be restored; I assumed that 
less than the whole system disk could/would help.

> what look to be a stack of several devops
> errors, and no ready path for recovery, and fossil-grade software
> versions?  Ah, fun times.
>
> NFS can throw odd errors [...]
>
> And unfortunately, the messages displayed by NFS don't always line up
> with reality; [...]
>
> See if SET WATCH FILE /CLASS=MAJOR or such might show which file is
> being accessed around the error.  Might have to start the NFS server
> directly to get messages from that, though.  [...]

As described in the base note, I figured that NFS was not the source of 
the problem, but the (first) victim.

> [...]
>
> Or upgrade this configuration to OpenVMS VAX V7.3 and "then-current"
> TCP/IP, which'll address a number of latent issues in this configuration.

As mentioned in a base note, upgrading is not an option. You would be 
surprised what applications are certified with exactly the software 
stack available on this (development) machine.

Anyway, as already stated elsewhere in this thread, portmapper wasn't 
the source of the problem either, it also was a victim.

Hans.