[Bioclusters] mpiblast error ObjMgrNextAvailEntityID failed

[Mike Muratet]

Greetings

This question appears to have come up before on these lists, but I don't 
see an answer. I recall we've posted about this problem before but I 
don't think we ever got the problem resovled. I created two databases 
using the commands below from the Ensembl human and mouse databases. I 
tried to run a query with:

setenv data_root /scratch/uahmam
setenv blast_root /opt/asn/apps/mpiBlast_1.3.0/latest/bin

setenv MPI_REQUEST_MAX 128000
setenv MPI_TYPE_MAX 128000

echo "blast_root is "$blast_root
echo "data_root is "$data_root
cd /scratch/uahmam

time /usr/bin/mpirun -np 8 $blast_root/mpiblast --remove  -p blastn -d 
Homo_sapiens.NCBI35.nov.cdna.fa -i novel_v11.fa -W 9 -r 1 -q -1 -G 1 -E 3 
-o novel_v_Hsen.blast

where my .ncbirc file is as below:

[mpiBLAST]
Shared=/scratch/uahmam
Local=/scratch/uahmam

[NCBI]
Data= /opt/asn/apps/ncbi_6.1/data

[BLAST]
BLASTDB=/scratch/uahmam
BLATMAT=/opt/asn/apps/ncbi_6.1/data


mpiformatdb --nfrags=6 -t enHs -i Homo_sapiens.NCBI35.nov.cdna.fa 
-p F -n enHs
mpiformatdb --nfrags=6 -t enMm -i Mus_musculus.NCBIM34.nov.cdna.fa 
-p F -n enMm

I get nothing but a million+ lines of:

ObjMgrNextAvailEntityID failed with idx 2048

It fills up my disk quota and crashes so I get nothing else.

I have found this line in the source code, but I can't figure out what's 
going on from that.

mpiblast seems to work OK for other databases/queries.

Can anybody offer any suggestions?

Thanks

Mike