[Info-vax] What is MDMS and how do I fix this error?

[Bob]

On Thursday, May 2, 2019 at 3:29:34 PM UTC-5, Stephen Hoffman wrote:
> On 2019-05-02 19:10:43 +0000, Bob said:
> 
> > I recently inherited an Alpha VMS 7.3-2 cluster from a former colleague.
> 
> Is this a production server, or is this a hobbyist server?   I'd assume 
> the former, but there's been all sorts of old gear passed around among 
> hobbyists.
> 
> Are there plans to support and keep this configuration current, or will 
> this configuration remain static on this now-unsupported OpenVMS 
> version?
> 
> VSI OpenVMS Alpha V8.4-2L2 is the current release, and support is 
> available on "recent" Alpha systems.
> 
> HPE OpenVMS I64 will remain supported for a little less than two years, 
> and HPE does not offer support for any version of OpenVMS Alpha.
> 
> > The following message has been filling the OPERATOR.LOG on all nodes:
> 
> I'd consider rotating the operator logs daily.
> 
> Purge out the older logs, archiving as necessary.
> 
> > %%%%%%%%%%%  OPCOM   7-APR-2019 00:26:08.65  %%%%%%%%%%%    (from node 
> > QWERTY at  7-APR-2019 00:26:08.65)
> > Message from user SYSTEM on QWERTY
> > Event: Access Control Violation from: Node LOCAL:.QWERTY Session Control,
> >         at: 2019-04-07-00:26:08.649-05:00Iinf
> >         NSAP Address=49::00-01:AA-00-04-00-1E-04:21,
> >         Source=UIC = [0,0]MDMS$SERVER,
> >         Destination=name = MDMS$SERVER,
> >         Destination User="",
> >         Destination Account="",
> >         Node Name=LOCAL:.QWERTY
> >         eventUid   BD25C7C3-58CB-11E9-85E0-005056916A71
> >         entityUid  8416B8C5-A625-11E8-8064-AA0004000304
> >         streamUid  95300A07-A625-11E8-807A-AA0004000304
> 
> That's usually a failing attempt to use a DECnet proxy.  Looks like 
> part of the backups is failing.  Or is no longer being used.
> 
> DECnet Phase IV host 1.30 seems involved.
> 
> Object 21 is not one I recognize off-hand, but may well be exactly this usage.
> 
>    Table 1 Object Type Numbers
> 
>    Number     Mnemonic  Description
> 
>    0          Task      User program
>    1-16                 Reserved for HP use
>    17         FAL       File Access Listener for remote file and
>                         record access
>    18         HLD       Host loader for RSX-11S downline task
>                         loading requests
>    19         NML       Network Management Listener Object
>    20                   RSTS/E media transfer program (NETCPY)
>    21-22                Reserved for HP use
>    23         REMACP    Network terminal handler (host side)
>    24                   Network terminal handler (terminal side)
>    25         MIRROR    Loopback mirror
>    26         EVL       Event receiver
>    27         MAIL      OpenVMS Mail Utility
>    28                   Reserved for HP use
>    29         PHONE     OpenVMS Phone Utility and RSX-11M/M-PLUS
>                         Phone Utility
>    30-41                Reserved for HP use
>    42         CTERM     Network terminal handler
>    43-62                Reserved for HP use
>    63         DTR       DECnet Test Receiver object
>    64-127               Reserved for HP use
>    128-255              Reserved for customer use
> 
> 
> > These happen about every 30 minutes on each node in the cluster.
> 
> ~48 messages a day times the cluster size shouldn't be a particular 
> load, given daily log rotation and archiving.
> 
> > I'm not sure what MDMS is, but it seems to be connected with ABS / SLS.
> > 
> > Any suggestions on what the problem might be, or how to fix it?
> > Any one know what MDMS is?
> 
> http://h30266.www3.hpe.com/odl/axplp/sysman/abs45/ABS044_CMD_GD.PDF
> 
> 
> 
> 
> -- 
> Pure Personal Opinion | HoffmanLabs LLC

Hoff;

Thanks for the tips. It's been a while since I've messed with ABS et al.

Production at client's site.

Staying on this version for the moment. No support contract.

We already rotate logs. Maybe "filling up" was the wrong word, but it puts enough stuff in the log that a review might miss something important.

1.30 would be the node not part of this cluster. Makes sense.

This has been ongoing for a while (long before I was at this site) and I finally had some time to dig a bit. I think I'll try the MDMS shutdown / restart and see if that clears things up.

Thanks again;

Bob