[ensembl-dev] VEP BUG - corrupt storable string

Fri Sep 14 13:33:30 BST 2012

Fixed

Will

On 14 September 2012 11:54, Duarte Molha <Duarte.Molha at ogt.co.uk> wrote:
> Thank you Will.
>
> I also noticed a small bug in the config method on the variant_effect_prediction script.
> If you set a different cache folder in your config file from the default location, the it ignores the setting on the config file.
>
> This happens because you set the cache directory to the default location on line 408
>
>     # dir is where the cache and plugins live
>     $config->{dir} ||= join '/', ($ENV{'HOME'}, '.vep');
>
> And then when in reaches that argument on the read_config_from_file method if statement on line 987 it fails to set that argument because it was defined previously.
>
> I believe you have to put the read config option before you set the default cache location on your script:
>
> (original)
>
>     # dir is where the cache and plugins live
>     $config->{dir} ||= join '/', ($ENV{'HOME'}, '.vep');
>
>     # dir gets set to the specific cache directory later on, so take a copy to use
>     # when configuring plugins
>
>     $config->{toplevel_dir} = $config->{dir};
>
>     # ini file?
>     my $ini_file = $config->{dir}.'/vep.ini';
>
>     if(-e $ini_file) {
>         read_config_from_file($config, $ini_file);
>     }
>
>     # config file?
>     if(defined $config->{config}) {
>         read_config_from_file($config, $config->{config});
>     }
>
> (mod)
>
>     # config file?
>     if(defined $config->{config}) {
>         read_config_from_file($config, $config->{config});
>     }
>
>     # dir is where the cache and plugins live
>     $config->{dir} ||= join '/', ($ENV{'HOME'}, '.vep');
>
>     # dir gets set to the specific cache directory later on, so take a copy to use
>     # when configuring plugins
>
>     $config->{toplevel_dir} = $config->{dir};
>
>     # ini file?
>     my $ini_file = $config->{dir}.'/vep.ini';
>
>     if(-e $ini_file) {
>         read_config_from_file($config, $ini_file);
>     }
>
> Best regards
>
> Duarte
>
> -----Original Message-----
> From: dev-bounces at ensembl.org [mailto:dev-bounces at ensembl.org] On Behalf Of Will McLaren
> Sent: 14 September 2012 10:04
> To: Ensembl developers list
> Subject: Re: [ensembl-dev] VEP BUG - corrupt storable string
>
> Hi Duarte,
>
> I've committed a fix to the CVS tree for this, hopefully this will prevent the problems you were having.
>
> Thanks
>
> Will
>
> On 10 September 2012 18:04, Duarte Molha <Duarte.Molha at ogt.co.uk> wrote:
>> Thank you...
>> Sorry for insisting, but because you had asked me for more details
>> about the bug I wanted to make sure I had provided you will all the
>> information you required. I did not mean to rush you guys :-)
>>
>> Cheers
>>
>> Duarte
>>
>> -----Original Message-----
>> From: dev-bounces at ensembl.org [mailto:dev-bounces at ensembl.org] On
>> Behalf Of Will McLaren
>> Sent: 10 September 2012 16:20
>> To: Ensembl developers list
>> Subject: Re: [ensembl-dev] VEP BUG - corrupt storable string
>>
>> Hi Duarte,
>>
>> We are looking into this - rest assured we'll get back to you if we track down the cause of the errors.
>>
>> Will
>>
>> On 10 September 2012 15:52, Duarte Molha <Duarte.Molha at ogt.co.uk> wrote:
>>> Any news regarding these bugs?
>>>
>>> Cheers
>>>
>>> Duarte
>>>
>>> -----Original Message-----
>>> From: dev-bounces at ensembl.org [mailto:dev-bounces at ensembl.org] On
>>> Behalf Of Duarte Molha
>>> Sent: 06 September 2012 10:48
>>> To: Ensembl developers list
>>> Subject: Re: [ensembl-dev] VEP BUG - corrupt storable string
>>>
>>> My apologies Will, I know should have done it... but now the script is not giving me this error anymore so I would not be surprised if it was user error (me) in this case.
>>>
>>> In attachment is a config file I am using to annotate a large VCF
>>> file
>>>
>>> The two only recurrent bugs I am still getting have to do with the
>>> fork parameter and with memory not getting released from the script
>>>
>>> 1) when running with no forks, the script runs fine but it appears that memory does not get deallocated from the script so all RAM and cache gets used and eventually the script dies.
>>> 2) when using the fork parameter, the script eventually crashes, but
>>> I do not believe it to be related to filling the RAM like the
>>> previous case. The script just crashes and outputs a bunch of garbled
>>> text
>>>
>>> I hope this is more usefull than my previous bug report.
>>>
>>> Best regards
>>>
>>> Duarte
>>>
>>> -----Original Message-----
>>> From: dev-bounces at ensembl.org [mailto:dev-bounces at ensembl.org] On
>>> Behalf Of Will McLaren
>>> Sent: 06 September 2012 09:51
>>> To: Ensembl developers list
>>> Subject: Re: [ensembl-dev] VEP BUG - corrupt storable string
>>>
>>> Hi Duarte,
>>>
>>> For this, and any future bug reports, it would be really helpful if you could relay the conditions that give rise to this error i.e. the input, options used etc.
>>>
>>> Thanks
>>>
>>> Will
>>>
>>> On 4 September 2012 15:59, Duarte Molha <Duarte.Molha at ogt.co.uk> wrote:
>>>> I would like to report another VEP bug that has been killing my script:
>>>>
>>>>
>>>>
>>>> Corrupted storable string (binary v2.7) at ../../lib/Storable.pm
>>>> (autosplit into ../../lib/auto/Storable/thaw.al) line 415, <CHILD>
>>>> line 301427, at
>>>> /NGS_Tools/bin/VEP_26_testing/Bio/EnsEMBL/Variation/Utils/VEP.pm
>>>> line
>>>> 970
>>>>
>>>>
>>>>
>>>>
>>>>
>>>> Duarte Molha
>>>>
>>>>
>>>>
>>>>
>>>> _______________________________________________
>>>> Dev mailing list    Dev at ensembl.org
>>>> List admin (including subscribe/unsubscribe):
>>>> http://lists.ensembl.org/mailman/listinfo/dev
>>>> Ensembl Blog: http://www.ensembl.info/
>>>>
>>>
>>> _______________________________________________
>>> Dev mailing list    Dev at ensembl.org
>>> List admin (including subscribe/unsubscribe):
>>> http://lists.ensembl.org/mailman/listinfo/dev
>>> Ensembl Blog: http://www.ensembl.info/
>>>
>>> _______________________________________________
>>> Dev mailing list    Dev at ensembl.org
>>> List admin (including subscribe/unsubscribe):
>>> http://lists.ensembl.org/mailman/listinfo/dev
>>> Ensembl Blog: http://www.ensembl.info/
>>
>> _______________________________________________
>> Dev mailing list    Dev at ensembl.org
>> List admin (including subscribe/unsubscribe):
>> http://lists.ensembl.org/mailman/listinfo/dev
>> Ensembl Blog: http://www.ensembl.info/
>>
>> _______________________________________________
>> Dev mailing list    Dev at ensembl.org
>> List admin (including subscribe/unsubscribe):
>> http://lists.ensembl.org/mailman/listinfo/dev
>> Ensembl Blog: http://www.ensembl.info/
>
> _______________________________________________
> Dev mailing list    Dev at ensembl.org
> List admin (including subscribe/unsubscribe): http://lists.ensembl.org/mailman/listinfo/dev
> Ensembl Blog: http://www.ensembl.info/
>
> _______________________________________________
> Dev mailing list    Dev at ensembl.org
> List admin (including subscribe/unsubscribe): http://lists.ensembl.org/mailman/listinfo/dev
> Ensembl Blog: http://www.ensembl.info/