Need the answers

[babeo]

Could someone tell me what severe level and what can be done to solve the problem. Also, I need to understand the message.
Thanks

On database machine, I get constant message from /var/adm/messages file:
kernel nspacket: t_snd, No message currently available

kernel nconnect: t_rcvconnect, An event requires attention
kernel engine 1, os pid 10842 online
kernel nspacket: t_snd, Is it serious?

On another machine, I get a deadlock process message:
Tue Oct 2 18:09:12: Your server command (family id #0, process id #149) was deadlocked with another process and hasbeen chosen as deadlock victim. Re-run your command.
Then I try to grep that ps, here it is:
myserver:/TVO/HISTORY>ps -ef|grep 149
root 149 1 0 Mar 02 ? 0:22 /usr/sbin/inetd -s
root 24430 149 0 Aug 19 ? 0:00 in.telnetd
root 8036 149 0 Sep 28 ? 0:00 in.telnetd
root 13338 149 0 Mar 30 ? 0:00 /usr/dt/bin/rpc.ttdbserverd
root 14018 149 0 Aug 29 ? 0:00 in.telnetd
root 13932 149 0 Aug 29 ? 0:00 in.telnetd
root 13926 149 0 Aug 29 ? 0:00 in.telnetd
root 27578 149 0 May 24 ? 0:00 rpc.metamhd
sysbaseowner 12075 13934 0 08:17:57 pts/6 0:00 grep 149
No message currently available

Mon Oct 1 11:54:30: Your server command (family id #0, process id #125) was deadlocked with another process and hasbeen chosen as deadlock victim. Re-run your command.

......
Thanks.

 

[jimbopalmer]

In Oracle (I am assuming Sybase is similar) a deadlock occurs if one user has Table A locked and requests a lock on Table B when another user already has Table B locked and now requests a lock on Table A. Both users will patiently wait FOREVER unless the database picks a victim.

There are a number of ways to prevent deadlocking, the most simplistic of which is always lock tables in alphabetical order. (in Oracle, only Updates and Deletes can lock a table) The one thing you can't give for your heart's desire is your heart. - Lois McMaster Bujold

 

[babeo]

Hi jimbopalmer

I think I misleading you. I understand the concept of deadlock, and for the database machine, I think the process t_snd is related to OS (not sure, that is why I need to find out), so it's UNIX problem (?!)

On the second machine, there is no database on it, so the deadlock processes happen are just purely related to UNIX only. (you may see the userid is 'sybaseowner', but it is for consistency of naming, but I am sure with you that there is not any database on this machine, but just a X25 collection data)

what I need to know is the severe level of these problems, if it is severely, then I need a solution to fix it.
Thanks