Kbase 11848: Diagnosing system error 1337 on VMS
| Autor |
Progress Software Corporation - Progress |
| Acesso |
Público |
| Publicação |
5/10/1998 |
|
Diagnosing system error 1337 on VMS
INTRODUCTION
------------
This Technical Support Knowledgebase entry explains how to find out the
relevant VMS error message associated with the following Progress error:
nssread--read error condition (number) (1337). The data dictionary gives
the following explanation - A Decnet error condition occurred, probably
due to a client process disconnecting. Check your ssdef system file for
the status given.
SOLUTION
--------
For Example:
nssread--read error condition 8444. (1337)
Make a note of the condition number - in this case 8444.
At the $ prompt on vms do the following:
1) $a=8444 (set up a symbol a=condition number)
2) $show symbol a (this will give us the value in decimal, hex and octal)
output in this case is:
A = 8444 Hex = 000020FC Octal = 00000020374
make a note of the hex value.
3) $exit %X20FC
output in this case is:
%SYSTEM-F-PATHLOST, path to network partner node lost.
This message means that the circuit between the local machine and one of
the remote machines has gone down - machine crashed, rebooted.
CONCLUSION
----------
The decnet error message will vary dependent on circumstances. Always
evaluate the condition number given in the progress error message.
Some other errors the conditon number may equate to, and some basic areas to
check are as follows:
"Remote node is no longer accepting connects"
This usually occurs when decnet on the remote or local node is
being shut down. Check in ncp for the node status.
eg: NCP> show exec status (this should be on if there are no
problems)
"Insufficient System Resources at Remote Node"
Check the number of logical links in use
eg: NCP> sho known links
Check the maximum number of logical links (MAX LINKS)
eg: NCP> sho exec char (MAX LINKS should be > than logical links.
"Network Partner Exited"
This is a "catchall". Communication with the remote node is possible
but the program on the remote node neither accepted or rejected the
connection within the Decnet timeout period (incoming and outgoing
timer). This could be due to the remote node being heavily loaded
in combination with LRPSIZE too small for ethernet packets. If this
is the case it has to use non-paged pool for I/O buffers, which is
slower.
Check the contents of netserver.log.
Check accounting.
Progress Software Technical Support Note # 11848