Errors were found while building depot

[FaizirizaL]

I found this error (refer to attached file)when install patch to server - HPUX11.11 .Is it the patch installed properly?

[melvyn burnard]

I doubt it, as you have a disc space error:

ERROR: "speksap1:/tmp/PHCO_28427/depot": This operation will exceed the absolute limit of disk space on 1 volumes.


Create more disc space, and try again

[Luk Vandenbussche]

Hi,

Check the diskspace in /tmp with bdf

you might have more space in /var/tmp
Try to execute it from here

[rariasn]

Hi,

Use another FS to generate depot, for sample, /var/adm/crash

- Move PHSS_XXXX to /var/adm/crash
- Move create_depot script to /var/adm/crash
- Remove /tmp/PHSS_*
- Remove /tmp/depot
- Create new depot in /var/adm/crash

rgs,

[FaizirizaL]

here is the bdf's from the server

Filesystem kbytes used avail %used Mounted on
/dev/vg00/lvol3 311296 309400 1896 99% /
/dev/vg00/lvol1 298928 55928 213104 21% /stand
/dev/vg00/lvol8 4194304 3771816 421992 90% /var
/dev/vg00/lvol7 1548288 1003952 540136 65% /usr
/dev/vg00/lvol4 524288 206880 314992 40% /tmp
/dev/vg00/lvol6 3072000 1503088 1556720 49% /opt
/dev/vg00/lvol5 106496 84904 21592 80% /home
/dev/vg04/lvol1 20480000 9664268 10656976 48% /appl
/dev/vg04/lvol2 516096 1229 482695 0% /scripts
/dev/vg04/lvol3 10240000 283174 9645728 3% /backup

as you can see here, there is 99% on / ...is that the prob comes from? how to solve this prob? on /var and /tmp look ok. what do you think?

what file inside the / can be trim ? how about the core file ?

[FaizirizaL]

Hi rariasn,

Use another FS to generate depot, for sample, /var/adm/crash

- Move PHSS_XXXX to /var/adm/crash
- Move create_depot script to /var/adm/crash
- Remove /tmp/PHSS_*
- Remove /tmp/depot
- Create new depot in /var/adm/crash


^ can you explain futher on this method? ^

[Isralyn M]

You need to clean up your root filesystem. Typical culprits are (1) core files, (2) a regular file being created under /dev, and (3) using the root file system as repository. With core files, there quite easy to pick up. Do a find on core files, copy the core file to tape and send to HP for fingerprinting. On the 2nd item, review the list of files under /dev. There shouldn't be any regular files created under /dev. Look at the file size and permission set. For the 3rd item, there should be a mount point for data repositories. If a directory is merely created without a mount point, they will definitely flood / and you'll end up with a system not being available because / is full. It is good sys admin practice to monitor /, /var and /tmp.

Regards,

Ira

[FaizirizaL]

hi isralyn m,

I need your futher assistance, pls provide me a step2 to solve this issue..
- can i reduce size for core file?

thnk you,

[rariasn]

Hi FaizirizaL,

"create_depot" procedure use /tmp directory for this commands:

- sh PHXX_nnnn (unshar file)

This genera the files:

PHXX_nnnn.depot
PHXX_nnnn.text

- and "swcopy" for create the depot:

swcopy -x enforce_dependencies=FALSE \
-x mount_all_filesystems=FALSE \
-x reinstall=TRUE \
-x write_remote_files=TRUE \
-d -s $PWD/$PATCH \* @ $DEPOT_name


Your output bdf "command":

/dev/vg04/lvol3 10240000 283174 9645728 3% /backup

# mkdir /backup/tmp
# cp -p /tmp/PHCO_28427 /backup/tmp
# cp -p /tmp/PHCO_31923 /backup/tmp
# cp -p /tmp/create_depot.sh

Remove /tmp/files

# rm -r /tmp/PHCO_28427
# rm -r /tmp/PHCO_31923

# cd /backup/tmp

# $PWD/create_depot.sh

The script create /backup/tmp/depot, for copy PHCO_28427.depot and PHCO_31925.depot

# swinstall -s /backup/tmp/depot&

rgs,

ran