IBM TSM server

[Miroslav Lachman]

claudiu vasadi wrote:
> On Sun, Jun 7, 2009 at 9:38 PM, Wojciech Puchar <
> wojtek at wojtek.tensor.gdynia.pl> wrote:
> 
> 
>>Jun  7 21:10:47 da1 kernel: ad6: detached
>>
>>>Jun  7 21:10:47 da1 kernel: g_vfs_done():ad6s1d[READ(offset=86540058624,
>>>length=16384)]error = 6
>>>Jun  7 21:10:47 da1 kernel: g_vfs_done():ad6s1e[READ(offset=63203147776,
>>>length=16384)]error = 6
>>>Jun  7 21:10:47 da1 kernel: g_vfs_done():ad6s1d[READ(offset=86539010048,
>>>length=16384)]error = 6
>>>
>>
>>
>>isn;t it trying to read past the end of disk?
>>
> 
> 
> Hmm.. I guess you are correct. The question is why is it doing this ?

It can be caused by wrong disk label (partitioning). Some partition made 
bigger then real media [disk].
You can see the label by command disklabel ad6s1 and then compare size & 
offset values with `diskinfo -v ad6`

> And on both drives ?
> 
> 
> Feels to me a problem for the developers. HDD's didn't give me one problem
> since the day the got installed. I'm not saing that they are very good HDD's
> (in fact they are quite cheap ones) but fail all of a sudden ? And leaving
> this asside, they dnt give one error on masive transfers and still have very
> good transfer speed. What I'm saing is that hdd failure could be a
> posibility of course, but a unlikely one at this point.
> 
> My problem is that I do not know what tehnik TSM server uses for creating
> those files because at some point it fails. Strainge thing is that it goes
> over 1G. First it creates the file and then it populates the file up until
> the given limit (25G in this case).
> 
> I will try something tomorow. I will again start the tsm server to add the
> space (the 25G free space) to the pool and will monitor the file size up
> until the OS crashes. I'm very curious what's the size of the file when the
> OS crashes.