| Title: | DIGITAL UNIX (FORMERLY KNOWN AS DEC OSF/1) |
| Notice: | Welcome to the Digital UNIX Conference |
| Moderator: | SMURF::DENHAM |
| Created: | Thu Mar 16 1995 |
| Last Modified: | Fri Jun 06 1997 |
| Last Successful Update: | Fri Jun 06 1997 |
| Number of topics: | 10068 |
| Total number of notes: | 35879 |
Hi We have a advfs mount problem with a A4100 with Digital UNIX V3.2G. It`s an A4100 with two CPU`s (5/300 & 5-6/300), two KZPSA (A10) and a SW500 with two HSZ40C (V2.7 with patches) and some rz28 & rz29 disks. If we want mount an advfs fileset with disks on the HSZ1 an error-messages occurs: adv_dom#test_fset on /mnt: I/O Error and the advfs fileset isn`t mounted! On the console (or messages or kern-log) appears the following message: advfs_logger: can`t find log end bs_bfdmn_active: can`t open domain ftx log, tag = 0xfffffff7.0x0: E_CANT_FIND_LOG_END (-1077) advfs_logger: pg # 119, is bad; truncating log No messages appears in the binary.errlog (uerf)! We can mount an advfs fileset with disks on the HSZ2 without any problems! We can mount an ufs filesys with disks on the HSZ1 without any problems too! We did change the KZPSA, the slots from the KZPSA`s, HSZ40, Cables, Trylinkadapter, and we did test it with only one disk in the SW500 (on the HSZ1-side). Any suggestions! Thanks Lucius [Posted by WWW Notes gateway]
| T.R | Title | User | Personal Name | Date | Lines |
|---|---|---|---|---|---|
| 9771.1 | KITCHE::schott | Eric R. Schott USG Product Management | Mon May 12 1997 04:28 | 20 | |
Hi
Something happened that the log is corrupt...this is not good...
a few questions:
- Are you using prestoserve?
- Did you have other I/O errors or a system crash prior
to having this problem?
recovery options:
- safest thing to do is recover from backups, as some kind of
corruption has occured.
- you can try mount -d, from single user mode, then
unmount and run the check programs to see if the the domain is ok...
Eric
| |||||
| 9771.2 | is a new sys | NNTPD::"[email protected]" | Lucius Schmid | Mon May 12 1997 08:35 | 14 |
Hi Eric No, we don't use prestoserve! And there are no other I/O errors! This is a new system! We had install Digital UNIX V3.2G and this problem occurs from the beginning! Sorry for the incomplete infos! Regards Lucius [Posted by WWW Notes gateway] | |||||
| 9771.3 | DECWET::MARTIN | Mon May 12 1997 16:25 | 31 | ||
So if you do: # mkfdmn /dev/r???? adv_dom # mkfset adv_dom test_fset # mount adv_dom#test_fset /mnt You get: adv_dom#test_fset on /mnt: I/O Error and the console messages shown in .0? I assume/hope that there isn't any data on one of these disks that you can't get from a backup. 'Cause I don't think you can get it now. My first guess is a bad disklabel. If the disk is incorrectly labelled as an rz28/rz29 instead of an HSZ disk, it'll think there are more blocks available than there actually are. However, the domain log should be near the beginning of the disk, not near the end. If you do an lseek() then a read/write at the last block of the raw device file, you should be able to tell if the disklabel is bad. I assume your references to HSZ1 and HSZ2 are referring to the two HSZ40Cs. If it's the same hardware, and it works on one but not the other, compare disklabels. If those check out OK, then try putting a UFS filesystem on the disk(s) you're having trouble with, and fill it with data, then see if you've got corrupt data in there anywhere. --Ken | |||||
| 9771.4 | kzpsa | NNTPD::"[email protected]" | Lucius Schmid | Tue May 13 1997 08:52 | 9 |
Hi Incredible but true. We replace the KZPSA controller the second time and now everything is OK!! Thanks for your answers Lucius [Posted by WWW Notes gateway] | |||||
| 9771.5 | Similar problem | KYOSS1::GREEN | Wed May 14 1997 11:44 | 3 | |
We had similar problem on 4100. It was Horse module.
There is a blitz about possible data corruption and Horse module.
| |||||