HP OpenVMS Systemsask the wizard |
The Question is: We are currently experience some network performance problem with "Decnet file copy". For one AlphaServer 1200, it took 8 mins to transfer 51 files. The same transfer on other server using the same source is less than a minute. We did not see any errors in ncp nor in any log. We ran the ncp "loop node" command against the problem server. With packet size of 1400 and 1000 packets, it completes in ~ 7 seconds. With packet size of 1401 and 1000 packets, it completes in ~ 3.5 mins. We haven't be able to find what cause the problem. Why packet size is a factor to the delay? Any suggestion on how to troubleshoot this problem. The Answer is : This situation most typically implies the existence of a network-level problem or network-hardware-level error, or potentially a process quota or host resource limitation. This could also be a problem within the connection or settings with to the network switch, or potentially with the packet fragmentation, with the host or switch jumbo frames support, or with the switch negotiation. Please contact the support center, as details of the current software and hardware network configuration, the current ECOs applied, and the values of the host-maintained network error counters (errors including packet retransmissions, collisions, back-offs, etc) will be required.
|