Hmm.

We have made a little progress on this; it appears to be related to the
fact the storage agent and the client use 127.0.0.1 to communicate on
the same host, regardless of any COMMMEthod SharedMem options if the
host O/S is not AIX. The  latest wisdom from the real unix Guru's here
say the number of open sockets against the same port seems to be causing
big slowdowns.

We can have upto 12 dsmc instances writing data through the storage
agent, along with upto 36 dsmc processes waiting for drives to become
available, all running on the same host. The backup data is large 100MB
- 2G+ files.

Has anyone experience similar issues on non-aix storage agent machines?


Matt.


-----Original Message-----
From: ADSM: Dist Stor Manager [mailto:ADSM-L AT VM.MARIST DOT EDU] On Behalf Of
Richard Sims
Sent: Tuesday, May 24, 2005 12:18 PM
To: ADSM-L AT VM.MARIST DOT EDU
Subject: Re: Poor LanFree performance / mixed LAN xfer??

Matthew -

Your posting doesn't specify your OS or TSM software levels, but
errno 232 is ECONNRESET in HP-UX, which is simply the "Connection
reset by peer". This is to say that the side reporting the errno did
not knowingly close the connection, so believes that it was closed by
the other side.

The meaning of the 131 is less clear. The message context would
suggest that it is a TSM internal code, which would mean an API
return code value, which would have to be "Internal system error".

The presence of the ANR0480W on the server end, with no accompanying
messages, would indicate that something in between the server and
client caused the loss of connection. So it would seem that you do
have communication link problems. Possibly, an intervening firewall
has an idle timeout value and dropped the TCP session. The poor
performance may be autonegotiation in the NIC and/or switch.

    Richard Sims

On May 24, 2005, at 6:15 AM, Warren, Matthew (Retail) wrote:

> Here is an excerpt from the dsmerror.log for last night, it appears we
> have some kind of comms issue..
>
> I havent been able to find errno 232 or 131 in /usr/include/
> errno.h, nor
> anything on IBM's site.
>
> We can match the ANS1005E messages to ANR0480 messages on the TSM
> server, indicating the client ended the session...
>
> 05/24/05   05:12:44 ANS1809W Session is lost; initializing session
> reopen procedure.
> 05/24/05   05:12:47 ANS1005E TCP/IP read error on socket = 9, errno =
> 232, reason : 'Connection reset by peer'.
> 05/24/05   05:12:47 ANS1809W Session is lost; initializing session
> reopen procedure.
> 05/24/05   05:12:48 ANS1809W Session is lost; initializing session
> reopen procedure.
> 05/24/05   05:12:50 ANS1005E TCP/IP read error on socket = 9, errno =
> 232, reason : 'Connection reset by peer'.
> ...
> 05/24/05   05:13:05 cuIdentifyResp: Error 131 reading IdentifyResp
> from
> server.


___________________________ Disclaimer Notice __________________________
This message and any attachments are confidential and should only be read by 
those to whom they are addressed. If you are not the intended recipient, please 
contact us, delete the message from your computer and destroy any copies. Any 
distribution or copying without our prior permission is prohibited.

Internet communications are not always secure and therefore Powergen Retail 
Limited does not accept legal responsibility for this message. The recipient is 
responsible for verifying its authenticity before acting on the contents. Any 
views or opinions presented are solely those of the author and do not 
necessarily represent those of Powergen Retail Limited. 

Registered addresses:

Powergen Retail Limited, Westwood Way, Westwood Business Park, Coventry, CV4 
8LG.
Registered in England and Wales No: 3407430

Telephone +44 (0) 2476 42 4000
Fax +44 (0) 2476 42 5432