[Bioclusters] NCBI blast with -a 2 option

Joe Landman bioclusters@bioinformatics.org
Wed, 23 Jun 2004 23:10:43 -0400

On Wed, 2004-06-23 at 22:32, Ungsik Yu wrote:
> Hi,
> We had similar problem.
> After setting environment variable LD_ASSUME_KERNEL=2.2.5,
> NCBI blast is running well with -a 2 option.

Now this is interesting.  The LD_ASSUME_KERNEL is used to force a
particular DSO choice, and hence a particular ABI (how the program talks
to the kernel services, size/shape of data structures among other
things) choice.  See http://people.redhat.com/drepper/assumekernel.html
for details.  The threading ABI in particular seems to be one of the
primary reasons for this.

Specifically the threading libraries have changed significantly, so if
they are being linked on an old enough RH box, they may be picking up
some deprecated ABIs.  This would be a bad thing.

Ok, looking at the binaries from NCBI, you might need to use this for a
while until they upgrade their build machine's OS.  I would like to hear
if anyone is running into this with either the static binary or the RPMs
that we supply at
http://downloads.scalableinformatics.com/downloads/ncbi/ .  I don't
expect this, but it would be useful to know.


Joseph Landman, Ph.D
Scalable Informatics LLC,
email: landman@scalableinformatics.com
web  : http://scalableinformatics.com
phone: +1 734 612 4615