[mira_talk] Re: Error while loading adsfacts during denovo assembly
- From: Surya Saha <ss2489@xxxxxxxxxxx>
- To: mira_talk@xxxxxxxxxxxxx
- Date: Wed, 31 Aug 2011 21:59:14 -0400
Hi Bastien,
Apologies for intruding on your vacation and the late response. I was
waiting to hear back from the sysadmin. You were right! The temporary
directory had only ~250GB of free space on the node. I am now rerunning the
job off a larger disk and it has gone past the 7 hr mark where it used to
crash earlier. Will get back if something comes up again. Please let me know
when you have the new version available with checks for write failures.
Thanks!
-Surya
On Mon, Aug 29, 2011 at 4:59 PM, Bastien Chevreux <bach@xxxxxxxxxxxx> wrote:
> On Aug 26, 2011, at 15:41 , Surya Saha wrote:
>
> I am running a denovo assembly of a bacterial genome having ~150X coverage
> (sampled from 2000X+ coverage dataset) and 7 million Illumina paired end
> 100bp reads on a 64GB RAM machine using MIRA 3.4.0. Here are my parameters
> [...]
> I am having to run it on a NFS but that cannot be the problem. MIRA stops
> due to an internal error ("Error while loading adsfacts, less facts in file
> than calculated earlier!"). It was using <30 GB or RAM at that point so
> memory is not an issue. Any ideas about what could be going on? The STDOUT
> is pasted below. Thanks!
>
>
> Hello Surya,
>
> sorry for not responding earlier, I am on vacation atm.
>
> After looking through the log (thank you very much for sending it) and
> checking a couple of things in the code, I believe that you may have run out
> of disk space a bit earlier and that MIRA did not check for that. Would that
> be possible?
>
> In the mean time I'm adding checks for write failures at the places I
> suspect it occurred to you (should put this kind of checks anyway in).
>
> Best,
> Bastien
>
>
Other related posts:
