BLAST jobs crashing - any ideas?

[Gary Williams]

We are seeing increasing numbers of BLAST searches failing like this:

---------------------------------------------------------------------------
Reference: Altschul, Stephen F., Thomas L. Madden, Alejandro A. Schaffer, 
Jinghui Zhang, Zheng Zhang, Webb Miller, and David J. Lipman (1997), 
"Gapped BLAST and PSI-BLAST: a new generation of protein database search
programs",  Nucleic Acids Res. 25:3389-3402.

Query= Z92844, 151798 bases, CBD5EE91 checksum.
         (151,798 letters)

Database: new
           312,486 sequences; 300,934,448 total letters

Searching.......................................
---------------------------------------------------------------------------

With no further output or error messages.
It exits with a status value of 1.

We are running BLAST on SPARC and Intel machines with Solaris.
We have large amounts of swap space available.
(For example: Memory: 5120M real, 153M free, 322M swap in use, 12G swap free)

The query sequence size and database size of some typical failed searches
are as follows:

Query		Database
-----		--------
129,429		2,950,329,007
193,974		2,950,329,007
151,798		  300,934,448
151,798		2,950,329,007
183,521		2,950,329,007
151,798		  300,934,448
151,798		2,950,329,007
 14,931		   87,398,758
 30,000		   87,398,758
100,000		1,607,027,821
100,000		   87,398,758
283,438		  105,694,337
104,147		2,507,032,618

Is anyone else seeing this sort of thing?

Does anyone have any idaes as to the cause or cure?

Thanks,
Gary

Gary Williams               Tel: +44 1223 494522  Fax: +44 1223 494512
mailto:G.Williams at hgmp.mrc.ac.uk            http://www.hgmp.mrc.ac.uk/
Bioinformatics,MRC HGMP Resource Centre,Hinxton,Cambridge, CB10 1SB,UK

---