Hi,
Today I started getting errors when trying to read or write some
files on 2 filesets within the one ADVFS domain.
The machine is a dec 3000/600
running OSF 3.2C
the root, usr & var are advfs with users_domain running on top of LSM
the disk layout is as follows
# df
Filesystem 512-blocks Used Avail Capacity Mounted on
root_domain#root 262144 101422 142688 42% /
usr_domain#usr 4110480 2073346 1977568 51% /usr
var_domain#var 1657632 93544 1549440 6% /var
users_domain#users 6144000 376870 5767130 6% /usr/users
users_domain#tmp 110592 22988 87604 21% /tmp
with the users_domain the one with problems and its a RZ29b type disk.
The problem is as follows.
Doing :-
#cat /tmp/tt
cat: /tmp/tt : I/O error
or cat /usr/users/jzentvel/.sh_history
cat: /usr/users/jzentvel/.sh_history : I/O error
this can also occur when trying to create files or copy files in either of
the filesets but not always.
uerf -R shows no errors but in the kern.log file has the following sort of
errors,
Feb 13 13:19:50 sharky vmunix: Memory error corrected by processor
Feb 13 13:19:50 sharky vmunix: biu_stat = 0000000000002340
Feb 13 13:19:50 sharky vmunix: biu_addr = 0000000000108018
Feb 13 13:19:50 sharky vmunix: dc_stat = 0000000000000007
Feb 13 13:19:50 sharky vmunix: fill_syndrome = 000000000000004a
Feb 13 13:19:50 sharky vmunix: fill_addr = 0000000001014410
Feb 13 13:19:50 sharky vmunix: bc_tag = 0000000000402a12
Feb 13 13:19:50 sharky vmunix: ident = 0
Feb 13 13:21:59 sharky vmunix: advfs I/O error: setId 0x30d8dfc5.000109d8.2.8001
tag 0x0000000a.82c5u page 0
Feb 13 13:21:59 sharky vmunix: vd 1 blk 7577840 blkCnt 128
Feb 13 13:22:00 sharky vmunix: write error = 6
Feb 13 13:22:00 sharky vmunix: advfs I/O error: setId 0x30d8dfc5.000109d8.2.8001
tag 0x0000000a.82c5u page 8
Feb 13 13:22:00 sharky vmunix: vd 1 blk 7577968 blkCnt 128
Feb 13 13:22:00 sharky vmunix: write error = 6
Feb 13 13:22:00 sharky vmunix: advfs I/O error: setId 0x30d8dfc5.000109d8.2.8001
tag 0x0000000a.82c5u page 16
Feb 13 13:22:00 sharky vmunix: vd 1 blk 7578096 blkCnt 128
Feb 13 13:22:00 sharky vmunix: write error = 6
Feb 13 13:22:00 sharky vmunix: advfs I/O error: setId 0x30d8dfc5.000109d8.2.8001
tag 0x0000000a.82c5u page 24
Feb 13 13:22:00 sharky vmunix: vd 1 blk 7578224 blkCnt 128
Feb 13 13:22:01 sharky vmunix: write error = 6
Feb 13 13:22:01 sharky vmunix: advfs I/O error: setId 0x30d8dfc5.000109d8.2.8001
tag 0x0000000a.82c5u page 32
Note the Memory error which is the first entry in the kern.log file for at least
the last 5 days, but im not sure if it caused the problem or is unrelated.
I have logged the call with DEC & we have tried running
vchkdir -f /usr/users
vchkdir -d /usr/users
msfsck -t users_domain
We ran this several times & then did a reboot.
when shutting down the system, the machine did a panic with the following error
Feb 13 16:53:03 sharky vmunix: ADVFS EXCEPTION
Feb 13 16:53:03 sharky vmunix: Module = 36, Line = 3648
Feb 13 16:53:03 sharky vmunix:
Feb 13 16:53:03 sharky vmunix: panic (cpu 0):
Feb 13 16:53:03 sharky vmunix: syncing disks... done
But on reboot the system mounted /usr/users & /tmp ok but with the same errors
occuring again.
Can anybody offer any advice on what to try to correct the problem with ADVFS
or would I be better of deleting the domain & recreating it from scratch.
I will let the list know what the result is.
Thanks,
--
John Zentveld Internet: jzentvel_at_scu.edu.au
UNIX VAX/VMS System Administrator Room: B2.14
Southern Cross University Tel: (066) 203859
Lismore NSW Australia Fax: (066) 203033
Received on Tue Feb 13 1996 - 07:58:52 NZDT