DEV_WAIT STATES

[MichelleButler]

Hello, I wonder if anybody recognises the following situation. I am an IT Manager for a small company, we have an informix/aix server version 4.2. I have 10 printers connected, 7 printers are connected to a 16 port adapter and the remaining printers are on print servers. I am getting constant dev_wait states on the printers connected to the 16 port adapter. When everybody starts printer one printer will go down, only after when I check the handing shaking and bring the queue up again then another printer goes down. I have check the var/spool and I feel that I have enough space. I do not have problems with the printer connected to the printer server. Any suggestions.

Michelle

 

[aixqueen]

You can try checking the cables for the printers...the flow control may be incorrect for the cable and the type of printers you have...You can also check the Timeout set for the printers...If the line is really really busy, it will timeout one and then it will go down.

Just a question....what speeds do you have the printers at? baud rates.. flow control (try for hardware if you can)
Since the 16 port device is a composite of all speeds...It is possible if all were set really high, and were really really busy they would go down...

DEV_WAIT: Printer off-line, out of paper, wires bad,etc. Will go to DOWN after
time-out when CTS low. CHECK CABLE. This can also occur when printer has
jammed, loose cable, or bad cable, or bad flow control.

Timeout of Printers, what to do?

If the queue is defined with standard processing (backend
is rembak), adding -T60 to the backend line in /etc/qonfig may help. If the
queue is remote with local filtering, then -T60 can be put into the rembak flags
section of /usr/lib/lpd/pio/etc/piorlfb. If the queue is defined using the HP
Jetdirect software (piojetd backend), then -T60 doesn't apply.

lp0:
host = hostname
s_statfilter = /usr/lpd/aixshort
l_statfilter = /usr/lpd/aixlong
rq = queuename
device = dlp0

dlp0:
backend = /usr/lpd/rembak -T99

cp /usr/lib/lpd/pio/etc/piorlfb
/usr/lib/lpd/pio/etc/piorlfb.orig
# vi /usr/lib/lpd/pio/etc/piorlfb
typeset piorlfb_rbflags="-T99" #rembak flags

-------------------------------------------------------------
NO guarantee wakeup down printer daemon
Can be Added to cron to reset down print ques every so many minutes. Probably
best to do every 1/2 hours
-----------------------
#!/bin/ksh
#
#
THISHOST=`uname -n`
lpstat|grep DOWN|awk '{print $1}' >> /tmp/lpstatdown.log
if test -s /tmp/lpstatdown.log
then
for i in `cat /tmp/lpstatdown.log`
do
enable $i
done
echo "The date/time is `date` " >> /tmp/lpstatdown.log
echo "\nThese queues were DOWN and have been enabled." >>
/tmp/lpstatdown.log
mail -s &quot;$THISHOST Printer Status&quot; root@localhost < /tmp/lpstatdown.log
sleep 5
rm -f /tmp/lpstatdown.log
fi

 

[AIXSPadmin]

I agree with AIXQueen that it is a flow control problem. It has been about 3 years since I worked with printer connected to an AIX server, but prior to that I had 4.5 years of working with AIX server connected printers. We had the same problems at times when new printers were addes. Change the flow control until you find the correct one. I cannot tell you which one it is since it has been a few years, though we had Kyocera, IBM, HP, and Mannesmann Tally printers connected.