Thanks for the information/confirmation. I sure hope IBM will recognize and
fix this. Since we've only seen/know of 2-occurances, this is mostly an
annoyance.

Sounds like it is related to the problems we experienced when we tried to
do replication with Oracle backups. We run multiple RMAN processes
simultaneously and the replication statements kept breaking them so we
stopped replicating Oracle nodes.

Zoltan Forray
Spectrum Protect (p.k.a. TSM) Software & Hardware Administrator
VMware Administrator
Xymon Administrator
VCU Computer Center
zforray AT vcu DOT edu - 804-828-4807
Don't be a phishing victim - VCU and other reputable organizations will
never use email to request that you reply with your password, social
security number or confidential personal information. For more details
visit https://phishing.vcu.edu

On Jan 31, 2018 7:47 AM, "De Kezel, Frederik" <
frederik.dekezel AT arcelormittal DOT com> wrote:

> FYI.
>
> I have seen this problem on numerous occasions and with various client
> versions (Version 7, Release 1, Level 2.2 -> Version 8, Release 1, Level
> 2.0).
> Our analysis is that it is indeed related to replication (we have two
> servers that cross-replicate to each other).
> And to the fact that we often perform multiple backup sessions in parallel
> (SQL agent jobs that kick off dsmc.exe per database; we don't use TDP).
>
> It seems that each session wants to write replication failover values to
> the options file.
> And when multiple sessions try to do that in parallel, they collide with
> each other and end up completely wrecking the options file.
>
> Workaround we have implemented: allow TSM to write failover values to the
> options file just once and then mark it readonly.
> Subsequent sessions can't write to it anymore and hence can't destroy the
> options file.
> Only drawback: each TSM session now causes the following alert to be
> written to dsmerror.log:
>
> ANS4058I A write failure occurred while attempting to save node
> replication failover values to the options file.
>
> We know what causes these alerts and choose to ignore them.
> The backups themselves are not affected.
>
> Rgds,
>
> Frederik de Kezel | Industrial Windows Server Infrastructure
> ArcelorMittal Gent
>
> Industrial Automation & Modelling (IAM)
> John Kennedylaan 51 | 9042 Gent | Belgium
> -----Original Message-----
> From: ADSM: Dist Stor Manager [mailto:ADSM-L AT VM.MARIST DOT EDU] On Behalf 
> Of
> Zoltan Forray
> Sent: donderdag 25 januari 2018 16:05
> To: ADSM-L AT VM.MARIST DOT EDU
> Subject: [ADSM-L] dsm.opt file being corrupted/blanked
>
> **This Message originated from a Non-ArcelorMittal source**
>
>
> Folks,
>
> Wondering if anyone has seen this issue. The first time I saw it a few
> months back, I ignored it as a one-time issue.  But it just happened again
> so I am digging deeper.
>
> This time (don't recall what the OS/client version was for the first
> incident) it is a Windows 2008 R2 server with BA 7.1.0.0 (yes I know -
> 0-releases are always problematic).
>
> OS tech noticed the backups failing and went to investigate.  Trying to
> start the CLI - it hung.  Checked the dsm.opt file and found it
> blank/empty? It has been backing up for a long, long time with no reason to
> make any changes to the dsm.opt file.
>
> AFAIK, the only thing that should be mucking with the dsm.opt file, aside
> from manually editing it, is replication.  So, is this replication related
> or a client bug?  Did some quick Googling without any luck on a known
> problem/fix.
>
> --
> *Zoltan Forray*
> Spectrum Protect (p.k.a. TSM) Software & Hardware Administrator Xymon
> Monitor Administrator VMware Administrator Virginia Commonwealth University
> UCC/Office of Technology Services www.ucc.vcu.edu zforray AT vcu DOT edu -
> 804-828-4807 Don't be a phishing victim - VCU and other reputable
> organizations will never use email to request that you reply with your
> password, social security number or confidential personal information. For
> more details visit http://phishing.vcu.edu/
>