[maker-devel] Contigs failing - could not make datastore directory

[Felix Bemm]

Do you have a changelog for 2.29b?

On 07.10.2013 16:33, Carson Holt wrote:
> You can also try the 2.29 beta if you want (on the website).  Also when
> MAKER exits, it tries to delete anything left behind in temporary folders,
> so is it possible that it is creating space on exit after filling up /tmp?

I will try to monitor that!

> Given that the issue seems to go away when you switch to another
> directory, it suggests that this might be the kind of thing that is
> specific to something about the current state of your your system.

I agree with that. Interestingly it happens with the only directory 
where I would not expect that.

> Thanks,
> Carson
>
>
> On 10/5/13 10:09 AM, "Felix Bemm" <felix.bemm at uni-wuerzburg.de> wrote:
>
>> Hi Carson,
>>
>> there is definitely no quota neither a full local tmp device. I changed
>> to another tmp device, which is unfortunately residing on nfs but now it
>> seems to work. I am using the latest version of maker (2.28). The
>> complete job is running on a single machine with 32 mpi threads.
>>
>> Bests
>> Felix
>>
>> Am 04.10.2013 20:21, schrieb Carson Holt:
>>> Hi Felix,
>>>
>>> There are other potential issues as well, for example a file quota limit
>>> set by your administrator or full local /tmp which will be unique to
>>> each
>>> node and cannot be queried directly from the root node on a cluster.
>>> On a
>>> cluster one node may also not have the NFS location mounted.  Or it
>>> could
>>> be system related, for example on ibrix NFS mounted systems you can get
>>> a
>>> weird filesystem issue with ghost files and directories when using MAKER
>>> that can neither be created nor deleted.  Try running the failed contigs
>>> in a different directory (even if it's on the same disk).  Also what
>>> version of MAKER are you using?
>>>
>>> --Carson
>>>
>>>
>>>
>>> On 10/4/13 3:10 AM, "Felix Bemm" <felix.bemm at uni-wuerzburg.de> wrote:
>>>
>>>> Hi,
>>>>
>>>> I have a problem with Contigs failing like this:
>>>>
>>>> examining contents of the fasta file and run log
>>>> ERROR: could not make datastore directory
>>>> --> rank=11, hostname=r5n04
>>>> ERROR: Failed while examining contents of the fasta file and run log
>>>> ERROR: Chunk failed at level:0, tier_type:0
>>>> FAILED CONTIG:MA_gen_v8.05_c31675
>>>>
>>>> There is enough space available on both the storage device and the tmp
>>>> device that is being used. The datastore log files looks like this:
>>>>
>>>> MA_gen_v8.05_c1149              FAILED
>>>> MA_gen_v8.05_c1153              FAILED
>>>> MA_gen_v8.05_c1154              FAILED
>>>> MA_gen_v8.05_c1155              FAILED
>>>> MA_gen_v8.05_c1156              FAILED
>>>>
>>>> Since there is no datastore for this contigs, I can not provide any
>>>> more
>>>> informationen. Approx. 1/3 of the contigs were annotated, after that
>>>> maker started to fail. It kept running for a while and finishing some
>>>> of
>>>> the contigs in the second or third try like this:
>>>>
>>>> MA_gen_v8.05_c4443		FAILED
>>>> MA_gen_v8.05_c4443		FAILED
>>>>
>>>> MA_gen_v8.05_c4443	MA_gen_v8.05M_datastore/A7/9B/MA_gen_v8.05_c4443/	STA
>>>> RT
>>>> ED
>>>> MA_gen_v8.05_c4443	MA_gen_v8.05M_datastore/A7/9B/MA_gen_v8.05_c4443/
>>>> FINISHED
>>>>
>>>> Any ideas what can cause this problems?
>>>>
>>>> Best regards
>>>> Felix
>>>>
>>>> --
>>>> Felix Bemm
>>>> Department of Bioinformatics
>>>> University of Würzburg, Germany
>>>> Tel: +49 931 - 31 83696
>>>> Fax: +49 931 - 31 84552
>>>> felix.bemm at uni-wuerzburg.de
>>>>
>>>> _______________________________________________
>>>> maker-devel mailing list
>>>> maker-devel at box290.bluehost.com
>>>> http://box290.bluehost.com/mailman/listinfo/maker-devel_yandell-lab.org
>>
>>
>> --
>> Felix Bemm
>> Department of Bioinformatics
>> University of Würzburg, Germany
>> Tel: +49 931 - 31 83696
>> Fax: +49 931 - 31 84552
>> felix.bemm at uni-wuerzburg.de

-- 
Felix Bemm
Department of Bioinformatics
University of Würzburg, Germany
Tel: +49 931 - 31 83696
Fax: +49 931 - 31 84552
felix.bemm at uni-wuerzburg.de