We have had the same problem since upgrading from 3.6.9 (we went to 3.8.l -> 3.8.10 -> 3.8.13 -> 3.8.15).
We have a production solaris 9 server (8 x 336 Mhz cpu's .. it runs other daemons as well) which services approximately 40 lab printers spread over 5 campuses. When a large number of jobs get into the queues we see the printers idle for significant periods of time between jobs when they should be busy printing the jobs out. When this happens we see printers which have very little in their queue go just as slow as the printers with a lot in their queue. Travis Greg Trounson wrote: > Greetings, > > I have a production RedHat7.1 server running LPRng-3.7.4-22, serving > two HP 4MVs in a laboratory of 30 PCs. > > At peak times, there may be 30 people trying to print two jobs each > through the server, resulting in 60 jobs on the print queue. > When the number of jobs gets this high, the print server appears to > slow to a crawl. The printers both sit pretty much idle, leisurely > printing a page every minute or so. This morning, the job turnaround > time (time from when someone clicked 'print' to when the paper came > out of the printer) had reached 75 minutes(!). > > CPU usage on the server is minimal (load average below 1.0), memory > usage is about 70%, it doesn't touch the swap partition, and there is > plenty of free disk space on all partitions. > The hard drive is going full-steam though. It seems that it is > continuously reading/writing the /var/spool/lpd/lp directory, which > now consists of sixty or so 20KB print jobs. A check of > /proc/interrupts reveals that the hard drive is communicating at about > 100 interrupts per second. > > The system remains in this state (printing out about a page per minute > and thrashing the disk while even more jobs accumulate) until people > stop sending jobs and the queue eventually falls below perhaps fifty > jobs. The remaining jobs are then rapidly processed and the quese > empties in a couple of minutes. > > Is there a known limit to the number of jobs that LPRng can handle? > Does it attempt to somehow sort the spool files on the disk whenever a > new job arrives (this is the only reason I can think of that the hard > disk would be thrashing), and if so, why? > > In accordance with advice on this list, I have since added all the lab > machines to the server's /etc/hosts file. I'm not sure how much that > will help in this case, but it's sensible, nonetheless. > > thank you in advance for any help, > Greg > > > ----------------------------------------------------------------------------- > > YOU MUST BE A LIST MEMBER IN ORDER TO POST TO THE LPRNG MAILING LIST > The address you post from MUST be your subscription address > > If you need help, send email to [EMAIL PROTECTED] (or lprng-requests > or lprng-digest-requests) with the word 'help' in the body. For the > impatient, > to subscribe to a list with name LIST, send mail to [EMAIL PROTECTED] > with: | example: > subscribe LIST <mailaddr> | subscribe lprng-digest [EMAIL PROTECTED] > unsubscribe LIST <mailaddr> | unsubscribe lprng [EMAIL PROTECTED] > > If you have major problems, send email to [EMAIL PROTECTED] with > the word > LPRNGLIST in the SUBJECT line. > ----------------------------------------------------------------------------- > > ----------------------------------------------------------------------------- YOU MUST BE A LIST MEMBER IN ORDER TO POST TO THE LPRNG MAILING LIST The address you post from MUST be your subscription address If you need help, send email to [EMAIL PROTECTED] (or lprng-requests or lprng-digest-requests) with the word 'help' in the body. For the impatient, to subscribe to a list with name LIST, send mail to [EMAIL PROTECTED] with: | example: subscribe LIST <mailaddr> | subscribe lprng-digest [EMAIL PROTECTED] unsubscribe LIST <mailaddr> | unsubscribe lprng [EMAIL PROTECTED] If you have major problems, send email to [EMAIL PROTECTED] with the word LPRNGLIST in the SUBJECT line. -----------------------------------------------------------------------------
