We have an old GIS system running on unix 10.20. Our IT department have kinda washed their hands of it I'm having a problem where the system seems to but full. It had been working find for the last few years but now the drive seems to be full. It is only used for viewing old information so nothing has been added to it in years. I take it some sort of temp files are filling up the drive?. I'm a bit of a novice but have root access so should be able to try some solution if anyone has some?.
Cheers
Sorry probably should of put this on the Unix for dummies section
What kind of box is this a 9000/700 or 9000/800 (wkstation or server)? or is the E part of the server type ( E35 E55 ?)
I dont remember if 10.20 had model command
could you show us the output of bdf ?
and vgdisplay to start with?
Even if nothing is added, system logs continue to grow you know...
And the output of uname -a
Last edited by vbe; 10-09-2014 at 06:11 PM..
Reason: added last line...
It's a E9000/813, which I'm told is on its last legs but that will be another thread(looking to back it up or move it to a newer system if that's even possible). I'm not in the same office as the box(using mobaxterm to access it) so used uname -a (HP-UX gis B.10.20 E9000/813) to get the details on it.
Yeah I was thinking it is some sort of log file filling up just don't know what to remove. I'm back in work in the morning will post up the bdf and vgdisplay outputs
This could be several things, I'd be tempted to start with /var/adm. /var/log, /var/spool or /var/mail
Having had a brief re-read of the post - I'd say that the most likely culprit here is the mail file for the root user, the most common problem in my experience is output from cron jobs. In addition, you seem to have an issue with /home, so you should have a quick look in there.
Files to lookout for are wtmp, wmtpx and standard log files also worth looking in t mail directories for users with large mail files.
Regards
Dave
Last edited by gull04; 10-10-2014 at 08:40 AM..
Reason: Additional Information.
Yes lvol8 is full... and you have a D330 server... that means with optional 2 disk on FW-SCSI board or max 7 disks on standard internal SCSI... which seems to be your case...
My concern is that all is in the same VG, not cool at all in case of crash...
As I have no more HP boxes, I have to try to remember what 10.20 had and usabel on a D class, I would start by
Code:
swlist -l product |grep ONLINE*JFS
but I will have to try uppercase, lowercase or mix.. We are looking for the onlineJFS product that if you have will save you... having to go single user to do weird manipulations...
Only this can be done only when you have freeded some space, for you have some unused blocks of disks: Free PE 134 That a bit more than 400MB...
Your /var is too small even for a 10.20 D class, I always allocated a minimum of 750 and that was short... Because is /var there is all the patches also, and a 10.20 fully patched thats more than 750MB...
Your case its oracle I suppose that writes in /var/tmp ( or /var/opt/oracle )
My suggestion as I dont what you know about HP-UX and how familiar you are with it ( I staterd on HP-UX 8.04...that was a long time ago on a 822 and a 855...) is you use SAM:
type sam -> there is somewhere where it says in one menu "trim logs? " Go there and trim all the logs If it complains about nop space ( you must be root though...)
go to /var/adm/syslog/and do a
Code:
rm OLD*
And see if that helps
then go in /var/tmp and try to remove all unwanted old files you find
but /var/tmp might be a sybolink link to /tmp ...in which case it will not help...
Have a look in /var/spool also
But most probably /var/adm will be the highest consumer for here lies also all the patches...
So the idea is to birng /var to about 90 % so that you can extaned it to about 750-780 MB so you keep about 100-150MB spare for any other space realted issue where that little amount can make all the difference
Hey all,
What do you think mostly happened in the following situation?
I have a Red Hat 5.5 server. Someone, somehow, managed to get two .nfs000.... type files that totaled over a terabyte in size. I removed them and thought things were back to normal. Then I started getting complains from... (2 Replies)
Hi,
In order to have a sand box machine that I could use to test some system changes before going to production state, I'd like to duplicate a working system to a virtual one. Ideally, I'd like to manage to do it this way :
- Make a full system backup excluding the user file system (this... (7 Replies)
hello
Even though I am not out of inodes or of space, the /var/adm/messages shows messages:
file system full
I am doing now fcsk -m (400G) and I am still waiting to see the fragmentation results (should I add another option to df to have a faster output?)
Do you have any other hints... (6 Replies)
I have unix server with OS 5.8 ,,,
I tried ufsdump 0ua -f /dev/rmt/0 / to perform full system backup on tape but I failed could any one give a procedure for full system backup on solaris machine using tapes??? (1 Reply)
I am receving following Error message in /var/adm/messages
"NOTICE: alloc: /: file system full"
Disk space usage is as beklow:
df -k
$ Filesystem kbytes used avail capacity Mounted on
/dev/md/dsk/d10 76678257 56962561 18948914 76% /
/proc ... (8 Replies)
I read the sticky and thought of a script I use on a regular basis. Since unless you patch/upgrade the df command on solaris you have a very tought time teling how full the system truly is.
Output looks like
$ biggest.sh /tmp
Filesystem kbytes used avail capacity Mounted... (0 Replies)
Hi, I just started working with UNIX on an old semi-fossilized Sun workstation which I use to process LOTS of images,however, I just started to get an error message that the file system is full and then my shell tool or/and text editor freeze up. Help? (8 Replies)
Hi All,
There was a background process running on a Solaris 2.8 machine, and appeared to have filled all available disk-space. I done a killall, and upon re-booting found that the file system had filled up, and will not boot as normal as a result. For example, I'm getting
/usr/adm/messages: No... (8 Replies)
When I try to log in as root I get the following message
realloccg /: file system full sendmail :NO Queue:low on space (have 0,SMTP-DAEMON needs 101 in /var/spool/mqueue) What should I do? (1 Reply)
10-09-2014
