All,
Has anyone tried using "rsync" with Sharity? When I attempt to sync the mounts on two 
servers I receive the message:

mkstemp fpix/.a.txt.jkaar0 failed: Operation not supported on transport endpoint

I bet this can be solved through a configuration change.

Christian,
I have fixed the problem. As it turns out the setup process failed and a number of % 
variables remained in the config file. I have manually set them and the strange 
behaviour has disappeared. Apologies for not noticing this sooner.

David
> 
> From: Christian Starkjohann <[EMAIL PROTECTED]>
> Date: 2003/02/19 Wed AM 05:35:58 EST
> To: [EMAIL PROTECTED]
> CC: [EMAIL PROTECTED]
> Subject: Re: [Sharity-talk] File corruption on Solaris
> 
> On Tuesday, Feb 18, 2003, at 23:34 Europe/Vienna, David Andrews wrote:
> > Thanks Mike,
> >
> > I think this is similar. Here is an example from the Solaris side:
> > # sum -r /mount/AMM001
> > 38207   4257 /mount/AMM001
> > # cp /mount/AMM001 /tmp
> > # sum -r /tmp/AMM001
> > 38207   4257 /tmp/AMM001
> > # cp /tmp/AMM001 /mount/AMMxxx
> > # sum -r /mount/AMM001
> > 42616   4257 /mount/AMM001      (***!!!!!!!***)
> > # sum -r /tmp/AMM001
> > 38207   4257 /tmp/AMM001
> >
> > What has happened above is that the original checksum on the original 
> > file is reporting a false value (the checksum of a corruption). The 
> > file that's stored in /tmp is also corrupted in the same way (and 
> > therefore has the same checksum). When the filesystem is fooled with 
> > (by storing the already-corrupted /tmp file into AMMxxx) the 
> > corruption changes and therefore the original file appears to have 
> > changed.
> >
> > I wonder if the reported 42616 is the correct checksum. If so, it 
> > would match your results. Are your files very large? This file is only 
> > about 2MB.
> 
> It seems I have not received the mail you replied to. File corruption 
> is a very serious problem and we don't have an open bug report 
> involving such a problem. If anyone gets corrupted data through 
> Sharity, PLEASE REPORT THIS PROBLEM!
> 
> David: To find out what's going on, we should try to get a more 
> controlled environment. Can you please do the following:
> 
> 1. Get a test file which reproduces the problem.
> 2. Transfer the file to the server by other means (ftp, floppy disk or 
> whatever).
> 3. Read it through Sharity and compare with the original.
> 
> If the files don't match, please run GNU diff to see where the content 
> has changed. Or run at least 'cmp' (maybe with option '-l') to see the 
> offset of the corruption.
> 
> It is possible that the corruption only occurs with particular 
> programs. You can also try running cmp directly between the Sharity 
> mounted file and the original, or use 'cp' to copy the mounted file to 
> /tmp and compare it there. This may make a difference.
> 
> Regards, Christian.
> 
> PS: Is this with the NFS2 or NFS3 frontend? Does it make a difference 
> if you use the other?
> 
> -- 
> Dipl.-Ing. Christian Starkjohann
> Objective Development
> mailto:[EMAIL PROTECTED] | http://www.obdev.at/
> 
> 

_______________________________________________
Sharity-talk mailing list
[EMAIL PROTECTED]
To unsubscribe see http://at.obdev.at/mailman/listinfo/sharity-talk

Reply via email to