I have a problem with 4.3.2 and the print spooler. Sometimes it queues up a report and then the printer goes into dev_wait for no reason. You can see with "lpstat -t" that there are jobs queued up but the printer just decides to quit. When I start killing the jobs sitting in the queue it eradicates all jobs EXCEPT that very last one, which it pretends to delete but the job continues to show up as an existing job and the printer remains in this dev_wait state which prohibits me from doing anything except rebooting. Then I get the queue back. On systems with over sixty users this is not convenient so obviously I'm looking for a "fix" to this problem. I have a feeling that there is going to be a patch for this problem. Any ideas?<br>
Tek-Tips is the largest IT community on the Internet today!
Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!
-
Congratulations strongm on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!
dev_wait 1
- Thread starter minervino
- Start date
A couple of questions, probably obvious but bear with me:<br>
<br>
1) Have you tried stopping and starting the print spooler when you get these problems?<br>
2) How is the printer physically connected? Is it a direct connection? a network printer?<br>
<p>Mike Lacey<br><a href=mailto:Mike_Lacey@Cargill.Com>Mike_Lacey@Cargill.Com</a><br><a href= Cargill's Corporate Web Site</a><br>
<br>
1) Have you tried stopping and starting the print spooler when you get these problems?<br>
2) How is the printer physically connected? Is it a direct connection? a network printer?<br>
<p>Mike Lacey<br><a href=mailto:Mike_Lacey@Cargill.Com>Mike_Lacey@Cargill.Com</a><br><a href= Cargill's Corporate Web Site</a><br>
-
1
- #3
Minervino-<br>
<br>
It's been a while since I used AIX (a couple of jobs ago...), but I remember there being a few problems with the print service. IBM took it on therselves to replace good old "lp" scheduler with their own print system (seemingly proprietary, IIRC) which had a few problems.<br>
<br>
One of the problems was that if you had network printers attached, and one of the network printers went down, the whole print service would hang up. The network printer would show as "dev_wait" status. Printers would show as being online, jobs would queue, but nothing would print.<br>
<br>
The point of all that is that it may be another bug you're seeing. Maybe there's a fix from IBM available? However, I've also seen similar problems to the one you describe when running traditional "lp" print services.<br>
<br>
You kill all the jobs on a printer off, but the last one hangs and shows a message such as "cancelled: informing user". The "fix" for this is to stop the print service and restart it.<br>
<br>
So, maybe, until you can get a permanent fix, you could restart the printer subsystem on the AIX box? This way, you wouldn't have to wait for a chance to restart the whole server every time you had this problem.<br>
<br>
This is where my memory fails me, I'm afraid... IIRC, the sequence of events for stopping and restarting the print queue when we had the network printer problem was something like:<br>
<br>
* Run ps and identify the PID of the printer subsystem daemon.<br>
* Kill the daemon.<br>
* Run a "print subsystem daemon stop" command to stop the printer subsystem.<br>
* Run a "print subsystem daemon start" command to start it again. (I think you can do these last two steps using <FONT FACE=monospace>smit</font>.)<br>
<br>
Printing may then be OK.<br>
<br>
Sorry for being so vague, but I hope this at least points you in the right direction.
<br>
<br>
Let us know how you get on.<br>
<br>
Thanks again Mike for pointing this problem out in the General Unix Discussion area
<br>
It's been a while since I used AIX (a couple of jobs ago...), but I remember there being a few problems with the print service. IBM took it on therselves to replace good old "lp" scheduler with their own print system (seemingly proprietary, IIRC) which had a few problems.<br>
<br>
One of the problems was that if you had network printers attached, and one of the network printers went down, the whole print service would hang up. The network printer would show as "dev_wait" status. Printers would show as being online, jobs would queue, but nothing would print.<br>
<br>
The point of all that is that it may be another bug you're seeing. Maybe there's a fix from IBM available? However, I've also seen similar problems to the one you describe when running traditional "lp" print services.<br>
<br>
You kill all the jobs on a printer off, but the last one hangs and shows a message such as "cancelled: informing user". The "fix" for this is to stop the print service and restart it.<br>
<br>
So, maybe, until you can get a permanent fix, you could restart the printer subsystem on the AIX box? This way, you wouldn't have to wait for a chance to restart the whole server every time you had this problem.<br>
<br>
This is where my memory fails me, I'm afraid... IIRC, the sequence of events for stopping and restarting the print queue when we had the network printer problem was something like:<br>
<br>
* Run ps and identify the PID of the printer subsystem daemon.<br>
* Kill the daemon.<br>
* Run a "print subsystem daemon stop" command to stop the printer subsystem.<br>
* Run a "print subsystem daemon start" command to start it again. (I think you can do these last two steps using <FONT FACE=monospace>smit</font>.)<br>
<br>
Printing may then be OK.<br>
<br>
Sorry for being so vague, but I hope this at least points you in the right direction.
<br><br>
Let us know how you get on.<br>
<br>
Thanks again Mike for pointing this problem out in the General Unix Discussion area
- Thread starter
- #4
OK minervino,<br>
<br>
Is it a remote or local printer?<br>
what are the entries for this device in the /etc/qconfig file?<br>
How are you removing these jobs?<br>
Did you use stopsrc/startsrc ?<br>
Output of lslpp -l spooler ( i think not sure )
<br>
Is it a remote or local printer?<br>
what are the entries for this device in the /etc/qconfig file?<br>
How are you removing these jobs?<br>
Did you use stopsrc/startsrc ?<br>
Output of lslpp -l spooler ( i think not sure )
- Status