Our network problems are not quite as dire as my last message stated. We are experiencing intermittent network problems in what seems to be a random sampling of our images . Interactive sessions with affected images begin to drag and, when pinged, they seem to drop the first few packets. The dramatic network outages that we experienced earlier were due to a router image that was experiencing the aforementioned symptoms. When the router was screwed up, every other image suffered, with frequent network outages across the board.
The only effective remedy to these problems that I have found so far is to stop the network services, unload the qeth and qdio modules and to then restart the network. As I am using Red Hat 7.2, I issue the following commands "service network stop; rmmod qeth qdio; service network start". This is generally a quick procedure, and my SSH connections aren't usually broken by it. However, this is only a temporary fix, as many of our images have experinced these delays multiple times. We are also talking to IBM about this problem. I'm curious as to whether or not anyone on z/VM 4.2 is experiencing any similar issues, as all of the recent complaints about mysterious network issues seem to involve versions 4.3 and 4.4. >We are running a number of qdio qeth guest lans that are experencing >very poor response. While accessing these servers via putty SSH I will >drop connection often. Sometimes within 5 minute intervals. I when so >far as to try ping 1000 from an iptables guest lan server that >front-ends an apache webserver guest lan and this ping received >multiple multi-second responses (as high as 59 seconds) as well as a >48% packet loss 516 out of 1000 packets accepted. I have been following >this listserv and noticed Micheal Lambert's zVM 4.4 guestlan problems >thread and thought this may be related but we run zVM 4.3 planning to >upgrade to 4.4 shortly. Has anyone seen this type of response from >guestlan on zVM 4.3 for qeth, qdio? Any thoughts on how I should >troubleshoot?
