Hello Everyone! I had an issue I had posted a little while ago with the following message:
ANS1005E TCP/IP read error on socket = 5, errno = 131, reason : 'Connection reset by peer'. Here is what I have done so far. Changed mpthreading to no which was supports suggestion. I have also increased the TSM servers idletimeout to 90 minutes, which had been 60 minutes. Are there any suggestions as to what this parameter should be on a tsm server on the mainframe? There are no other messages in the error log on the clients other than the above message. I was wondering, for a SUN Solaris client, what the tcpwindowsize should be? I have looked in the performance tuning guide and it states at the beginning of the client section that it should be 64. Should this number be increased to improve backup/archives? When I looked in the activity log on the TSM server which is on os/390 2.10 at level 5.1.6.2, the only thing I could see was that a session ended for the client in question. Then seconds later another session started and within 2 minutes it ended and a majority of the time it backed up at least 6GB if not more, so I am going to assume that it just continued where it had left off and then finished its backup. From my point of view this appears to be a client issue because I cannot find anything in the activity log that states otherwise. Any suggestions/comments would be appreciated! Thanks! Joni Moyer Systems Programmer [EMAIL PROTECTED] (717)975-8338
