STAR core dump

[ymc]

With so little RAM, it doesn't make sense to use rna-star

[GenoMax]

Defining large amounts of swap does not help with lack of real memory. I concur with Devon about finding alternate hardware.

If you are working with NGS data then 512 MB of RAM (I really hope 0.5G is a typo) is not going to get you anywhere with any program.

STAR major memory requirements posted by Alex in a thread on STAR on the google group:
1. Genome + SA: ~9*GenomeLength bytes
2. SAindex: ~6*4^(--genomeSAindexNbases). By default --genomeSAindexNbases 14 and it takes 1.5GB of RAM.

[dpryan]

That's unlikely to be enough, unless you work on bacteria or something like that. You'll need to use a bigger computer.

[huzefa_mehta]

I compiled the libraries myself

I have about 0.5G of RAM, however have about 12G of swap virtual memory.

[dpryan]

In addition to what GenoMax wrote, how much RAM does the computer you're using have?

[GenoMax]

What OS is this? Did you compile STAR on this machine or download the binaries?

[huzefa_mehta]

I havent been able to run STAR successfully yet. It exits with a core dump message. Any clues?



Aug 29 20:52:35 ..... Started STAR run
Aug 29 20:52:35 ... Starting to generate Genome files

terminate called after throwing an instance of 'std::bad_alloc'
what(): std::bad_alloc