interesting 'rasplit' behavior

Carter Bullard carter at qosient.com
Tue Oct 24 22:38:07 EDT 2006


Hmmmm, I must have lost count of the number of suffix characters.
I'll try to get a fix in quickly.  By the way, thats a lot of output  
files?

Carter

On Oct 23, 2006, at 7:16 PM, MN wrote:

>
> Hi -
>
> Running rasplit (with the implied default "-M count 10000") on
> a large file:
>
> # ls -al /d/slurp2/test.bz2
> -r--r--r--  1 xfer xfer 990913849 Sep 17 18:00 /d/slurp2/test.bz2
> # /usr/local/src/argus-clients-3.0.0.rc.33/bin/rasplit -r /d/slurp2/ 
> test.bz2
>
>
> produces an output ordering that does not match the man page:
>
> 	When  rasplit  is using the default count mode or the
> 	size mode, the suffix is a group of letters 'aa', order
> 	by  file  name  produces  the  original  input  file.
> 	If rasplit will need to create more output files than
> 	are allowed by the default suffix strategy, more letters
> 	will be added, in order to accomodate the needed files.
>
> As it seems to do a wrap rather than an add:
>
> # ls -ltar
>
> (summarized, in order of file creation; note hte xzzzz -> xzaaa
> transition:)
>
> 	xaa
> 	xab
> 	...	(what you would expect here)
> 	xay
> 	xaz
> 	xba
> 	...
> 	xbz
> 	xca.xcz	(what you would expect here)
> 	... 	(what you would expect here)
> 	xya
> 	...	(what you would expect here)
> 	xyz
> 	xzzaa
> 	xzzab
> 	...	(what you would expect here)
> 	xzzaz
> 	xzzba
> 	...	(what you would expect here)
> 	xzzbz
> 	... ...	(what you would expect here)
> 	xzzza
> 	...	(what you would expect here)
> 	xzzzz
> 	xzaaa	(!!!!????) <------
> 	xzaab
> 	...
> 	xzamn	(last file)
>
> Which seems to violate both of the sentences from the man page.
>
>
> Thanks,
> - mike
>
>
> ps: I've also sent Carter an example of a
>
> 	ArgusAlert: ra[12980]: ArgusReadConnection: not Argus-2.0 data  
> stream.
>
> generating record, but can't share the record with the list.
>


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://pairlist1.pair.net/pipermail/argus/attachments/20061024/33dc7e63/attachment.html>


More information about the argus mailing list