File compression causes crash

[Hilda Mabhena]

attached is the crash data generated.

i'm trying to compress an export file. when it reaches a certain size, the system crashes and generates the attached log. please help. i need to write the file to tape but its just too big and will not be written.

[Steven Schweda]

What's "an export file"?

Trying to "compress" how?

How big is "too big"?

What does "will not be written" mean?

> _host_version: Digital UNIX V4.0D (Rev.
> 878); Mon May 24 07:59:32 GMT 1999

Not exactly the latest version, is it? Have
you installed any available patches?

[Hein van den Heuvel]

There is an IO error in that dump
That shouldn't tkae the system down, but I guess it did.
THis probably has nothing to do with the compression and everything with the IO.
Try :

1) the advfs command below to find the file which had the IO error

2) Copying the input file to /dev/nul to make sure it can be read

3) Try a different output device for the compressed data

4) split the export file, or if this is Oracle, re-export with a maximum output size causing it to split.

5) check out your swxcr and its drives.

Good luck!
Hein.

AdvFS I/O error:
Domain#Fileset: domain_data#data
Mounted on: /data
Volume: /dev/re0c
Tag: 0x00005b68.805c
Page: 170145
Block: 4804272
Block count: 128
Type of operation: Read
Error: 5
To obtain the name of the file on which
the error occurred, type the command:
/sbin/advfs/tag2name /data/.tags/23400

[Hilda Mabhena]

it gives reference to the export file which i'm trying to compress

[jim owens_1]

It appears you have a broken raid contoller or the storage behind it is bad:

> re_getdrive
> Cmd should always return good status
> Hard Error Detected
> Active XCR_COM at time of error
> xcr_logger: XCR_ERROR packet
> xcr_logger: cntrl 0 unit 0

Resulting in a read error on /dev/re0c

The kernel crashes while trying to recover from the failed read. A supported version of the kernel might not crash, but it won't get your data back from the bad device. You need to fix that.