[Bioclusters] ncbi blast

[Michael Gutteridge]

Hi

Apologies if I'm duplicating anything, but I couldn't find the fix in 
the list.  We've recently run into this problem with both 2.2.6 and 
2.2.9 on Linux 2.4.20-20.8smp (i686 i686 i386 GNU/Linux).  I ran 
through the basic checks indicated (libraries found, etc), though I 
haven't done memtest (this problem occurs on every host it's been run 
on in our cluster).

I've also run against the single small sequence, with the same results. 
  When I did an strace on the run, it seems to segv during an munmap 
call reading from segment 1 of the nt database (which is distressingly 
close to 1GB in size).

I don't know if others have seen this, but when I don't write to an 
output file (i.e. output to stdout), I see this message (which is 
actually on stderr, but doesn't seem to get written to error.log) in 
the output:

[blastall] ERROR: abc123|my: ObjMgrDelete: deleting locked element
[blastall] ERROR: abc123|my: ObjMgrDelete: deleting locked element
[blastall] ERROR: abc123|my: ObjMgrDelete: deleting locked element
[blastall] ERROR: abc123|my: ObjMgrDelete: deleting locked element

Could this be significant?  Or is this part of normal operations?

Someone was contacting NCBI- did that turn up anything useful?

Thanks

Michael


On Jun 23, 2004, at 11:09 AM, Joe Landman wrote:

>
>
> On Wed, 2004-06-23 at 13:36, Susan Chacko wrote:
>> On Jun 23, 2004, at 12:22 PM, Joe Landman wrote:
>
> [...]
>
>>>  If I build a static binary of blastall, would you be willing to give
>>> it a try?
>>
>> Yes, I'll definitely try it if you are willing to build a binary.
>
> http://downloads.scalableinformatics.com/downloads/ncbi/blastall.gz
>
> grab it and then
>
> 	gzip -d blastall.gz
> 	chmod +x blastall
> 	./blastall ....
>
> Static 2.2.9 binary, compiled with gcc 3.3.1, targeted at 586 CPUs
> (-mpentiumpro).  Please let me know if you see the problem recur.
>
>
> -- 
> Joseph Landman, Ph.D
> Scalable Informatics LLC,
> email: landman@scalableinformatics.com
> web  : http://scalableinformatics.com
> phone: +1 734 612 4615
>
> _______________________________________________
> Bioclusters maillist  -  Bioclusters@bioinformatics.org
> https://bioinformatics.org/mailman/listinfo/bioclusters
>