[ensembl-dev] Lock on index file

[Will McLaren]

Hi Konrad,

Sounds like an issue with Bio::DB::Fasta, which is part of the BioPerl
module set that we don't maintain.

A quick Google brought up this:

http://mailman.open-bio.org/pipermail/bioperl-l/2012-December/087835.html

which might be of some help; perhaps it depends on the index type that it
chooses to create (http://search.cpan.org/~cjfields/BioPerl/Bio/DB/Fasta.pm
).

It may also be a filesystem thing - I know NFS can have issues with
multiple processes accessing the same file.

Regards

Will

On 9 December 2014 at 18:40, Konrad Karczewski <konradk at broadinstitute.org>
wrote:

>   Hi dev team,
>
> I'm confused by an error I've been occasionally getting recently. I've
> created a sequence index using a test VEP run (and verified it's written
> and gets read by a test run without a problem), and then submit ~a few
> hundred VEP jobs to a cluster. A small subset of them come up with this
> error:
>
> Can't open sequence index file
> /humgen/atgu1/fs03/konradk/vep/gold//homo_sapiens/77_GRCh37/Homo_sapiens.GRCh37.75.dna.primary_assembly.fa.index:
> No locks available at
> /humgen/atgu1/fs03/konradk/ensembl-tools-release-77/scripts/variant_effect_predictor/Bio/DB/Fasta.pm
> line 527.
>
> I would have thought if the index existed and was readable, it wouldn't
> try to write it, but asking for a lock seems to suggest it is: any ideas
> what may be causing this? This happens on at least on VEP versions 77 and
> 78.
>
> Thanks!
>
>
> _______________________________________________
> Dev mailing list    Dev at ensembl.org
> Posting guidelines and subscribe/unsubscribe info:
> http://lists.ensembl.org/mailman/listinfo/dev
> Ensembl Blog: http://www.ensembl.info/
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.ensembl.org/pipermail/dev_ensembl.org/attachments/20141209/00e8da8f/attachment.html>