Even more out of office than normal so maybe I don't have the throughput to
process this but it sounds like it won't work then. If the new types aren't
going to be loaded than we cannot evolve the datatypes with new
functionality in new repositories.
Perhaps I am missing something, but in the abs
My understanding of the code is that tool shed dependencies (or local
dependencies) will not be available to tool shed datatypes (for
sniffing for instance). Sorry.
If you want to hack up your local instance to resolve dependencies
during the sniffing process that may be possible - my guess is you
I think the confusion over two different test methods has reached a
high level and Dave and I agreed previously that we should try to make
the API driven version the default during this release cycle - but it
is not our call :).
Greg - is there anything else you feel that needs to happen to at
lea
Does the current implementation really handle datatypes in reproducible
manner - if I have a repo which in revision 1 defines foo1 as a text
subtype, foo2 as a tabular type and foo3 as a new type in foo.py and then
in revision 2 foo1 is defined as a binary subtype , foo2 and foo3 disappear
and foo4
On Tue, Jul 22, 2014 at 7:51 AM, Peter Cock wrote:
> On Tue, Jul 22, 2014 at 1:15 PM, Eric Rasche wrote:
>> Hi Peter,
>>
>> On July 22, 2014 3:15:41 AM CDT, Peter Cock
>> wrote:
>>>
>>>Given how close you can get now for minimal effort,
>>>this seem unnecessary.
>>>
>>>http://blastedbio.blogspo
Running jobs as the "real" user is not available with the PBS job
runner - one has to use the DRMAA interface to submit jobs as the real
user.
I have created a Trello card to add this functionality:
https://trello.com/c/OddS8bMP
Would be happy to field pull requests to add this - because I doubt
Hey Eric,
This is not possible (easily anyway) in the latest release of blend4j.
It was on the TODO list though so I have added the functionality in
the following commit:
https://github.com/jmchilton/blend4j/commit/f92909fbda3616da09614b65810ebd86ce496b19
So instead of using GalaxyInstanceFactor
the time being.
Hmm... that is too bad. They should do that - seems like Android would
be a key platform for a web service client library. Oh well.
-John
>
> Thanks for being aware of this issue.
>
> Cheers,
> Eric
>
> 22.07.2014, 21:10, "John Chilton" :
>> Hey
Hello,
Thanks for the bug report. Is this on the main public server (usegalaxy.org)
or a local instance at the University of Iowa?
-John
On Wed, Jul 16, 2014 at 10:39 AM, Beck, Emily A
wrote:
>
> Hello,
>
> I have repeatedly gotten the following error message when attempting to
> use both the
Problem with automation is I could create dozens of templates over the
next several years and consume less time in aggregate than it would
take me to automate this. Nonetheless, there is a documentation
component here that is important so I did enough to document - if
someone wants to automate from
there but no values are there.
>
> Could you give me a use case, so I can mime it to my project ?
>
> Thanks in advance,
> --
> Olivia Doppelt-Azeroual
> CIB, Institut Pasteur, Paris
>
> Le 27/02/14 14:46, John Chilton a écrit :
>
>> I think you are right
Hello Julie,
Right now Galaxy can implicitly create collections by running a tool
that takes in a single input or a paired dataset over say a list of
inputs or a list of paired inputs but tools cannot explicitly create
dataset collections. I think Galaxy's current capabilities add a lot
of usefu
Hey Peter,
Opps sorry about that and thanks for the bug report. The tool shed
code should be fixed with
https://bitbucket.org/galaxy/galaxy-central/commits/38ba45d6ba5be65b3b743fc08739e16cd6e0ac8f
- it is in next-stable so I think the tool shed should pick up that
fix at next tool shed update.
-J
ing
> my tool once the fix is live on the Tool Shed?
>
> i.e. Was it a harmless warning?
>
> Peter
>
>
> On Wed, Jul 30, 2014 at 12:12 PM, John Chilton wrote:
>> Hey Peter,
>>
>> Opps sorry about that and thanks for the bug report. The tool shed
>> cod
, Jul 29, 2014 at 4:43 PM, David Kelly wrote:
> Hi John,
>
> Is there any documentation that explains how dataset collections work? I'd
> like to learn more about it.
>
> Thanks,
> David
>
>
> On Tue, Jul 29, 2014 at 3:20 PM, John Chilton wrote:
>>
>
I don't believe this is possible and I am aware of not active plans to
work on this. But if you are willing to hack on Galaxy I think you
could enforce arbitrary constraints pretty easily (if you know Python)
by modifying the validate_password function in
lib/galaxy/security/validate_user_input.py.
Thanks for tracking down the problem - it sounds like it is a Galaxy
bug then so I have created a Trello card
(https://trello.com/c/bNEKfOWR).
-John
On Wed, Jul 30, 2014 at 7:06 AM, Peter Cock wrote:
> On Wed, Jul 30, 2014 at 11:52 AM, Ulf Schaefer
> wrote:
>> Dear Nate, dear Peter
>>
>> Again
On Thu, Jul 31, 2014 at 12:29 PM, Peter Cock wrote:
> On Thu, Jul 31, 2014 at 5:21 PM, bjoern.gruen...@googlemail.com
> wrote:
>> Hi Peter,
>>
>>
>> 2014-07-31 10:57 GMT+02:00 Peter Cock :
>>
>>> Hi Dave,
>>>
>>> You are right that on closer inspection I've mixed tool_dependencies.xml
>>> and rep
On Fri, Aug 1, 2014 at 4:16 AM, Geert Vandeweyer
wrote:
> Hi,
>
> I'm running into issues with our galaxy installation regarding the
> new_file_path setting:
>
> 1/ clean up :
>
> Can I safely delete all contents in the tmp folder using a cronjob, or
> manually when galaxy is (not) runnin
Hmm...
The following should work for the last at least three releases I
think: ${__user__.api_keys[0].key}. Not sure why get/find are
unavailable on the User model class in the cheetah template now. I
think this version is a little better. Is this okay? If not I can
create a Trello card to investi
Nothing comes to mind immediately, can you post a link to your
job_conf.xml file?
What is your `cleanup_job` option set to in universe_wsgi.ini? Can you
try changing it to 'never' and see if the jobs continue?
-John
On Tue, Aug 5, 2014 at 5:54 PM, Langhorst, Brad wrote:
> I’m not aware of any r
This bug predates Galaxy's transition to mercurial - so I would
definitely encourage applying the patch by hand and restarting Galaxy.
-John
On Tue, Aug 5, 2014 at 9:13 PM, Dooley, Damion wrote:
> Does this apply to all past galaxy installs? I have an older galaxy site
> I've been wanting to p
Jobs run as the condor user? That is surprising - Galaxy isn't running
as the condor user is it?
Is your intention to have the jobs run as the 'galaxy' user or to run
as the LDAP authenticated end-user submitting the job?
If it is to run as the Galaxy user - I think it would be best to have
condo
Doubt this is job metrics related - fairly certain the problematic line is:
https://bitbucket.org/galaxy/galaxy-central/src/default/lib/galaxy/webapps/galaxy/controllers/dataset.py?at=default#cl-112
It tried to determine if a dataset id is encoded or not by hitting the
database. It could possible
City, IA 52242
> Lab: (319)335-3430
> ----------
> *From:* John Chilton [jmchil...@gmail.com]
> *Sent:* Tuesday, July 22, 2014 4:06 PM
> *To:* Beck, Emily A
> *Cc:* galaxy-...@bx.psu.edu
> *Subject:* Re: [galaxy-dev] Galaxy error
>
> Hello,
>
> Thanks for the bug report. Is thi
I have applied your patch here
(https://bitbucket.org/galaxy/galaxy-central/commits/19309418c86450b7f39691460eb430e100883cb4)
and one correction to James' previous comment - multi-page tools
actually do work through the API - this was required for multi-page
tool testing via the API.
-John
On Tue
Nate is right - full traceback would help. I am not sure it is the
only problem - but these outputs do not look right:
Galaxy isn't going to expand variables ($Project.input /
$Project.samples) in the from_work_dir attribute so those are going to
be taken as literal strings. That dire
Hey Peter,
When you are deleting and reinstalling the tool - can you confirm for
me that you are definitely deleting the tool and not just deactivating
it?
I don't know where to hack around from there - I think repositories
are tracked in the database (tool_shed_repository) table, in
shed_tool_co
Fun question! I have opened a pull request with my answer -
https://bitbucket.org/galaxy/galaxy-central/pull-request/457/allow-cheetah-tool-templates-to-reason/diff.
There are three different hacks you can use right now... here is a
diff against tools/filters/catWrapper.xml I was using the to test
You don't need a tool to upload - you need 'upload1' (i.e.
tools/data_source/upload.xml) and it cannot come from the tool shed -
this will make its id something other than upload1. You can make some
modifications to it, relocate it, etc... but that needs to be there in
some form.
In universe_wsgi.
along that chain to remove an entry from
> tool_shed_repository.
>
> The tool and its type dependency are attached.
>
> On 12/08/2014 16:43, John Chilton wrote:
>> Hey Peter,
>>
>> When you are deleting and reinstalling the tool - can you confirm for
>> me that
When Galaxy cannot find the extension/short name corresponding to a
datatype it treats the input as a 'data' the root datatype meaning
anything can connect to it. So I think the problem here is not so much
your tool format as your datatype - I don't think Galaxy can see the
lefse_res datatype. You
I am not sure what the problem is - more details would really help.
For instance - What version of Galaxy are you using? What repository
are you trying to install? Do other repositories work? Does the Galaxy
log have a more complete stack trace?
Any additional details would be most helpful.
Thank
When was the last time you updated Galaxy - we fixed a bug that would
be causing this behavior several releases ago.
Otherwise - additional information would help - like are you using the
DRMAA job runner or the newer specialized Slurm job runner and are
there any details in the Galaxy log that mi
Okay - I just checked out a clean Galaxy without modifications and I
don't see any problems with uploads. So this is likely something to do
with your setup - if I had to guess I would guess that your
job_conf.xml configuration is somehow incorrect - the job is getting
creating but nothing is runnin
So you want temp directories and working directories to sit on the
external drive but not Galaxy's files:
Lets say this external disk is mounted as /mnt/gxwork. I would create
two directories under that maybe /mnt/gxwork/tmp and
/mnt/gxwork/working.
Setting the working directory is the easier par
efse program, it
> generates output of the type lefse_res.
>
> So why
>
>
>
>
>
>label="Input 2"/>
>
>
>
> does not filter only those datasets?
>
> Thanks!!
>
> George
>
>
>
>
>
> On Wed, Aug 13, 2014 at 1:33 P
I am using Galaxy at an all out
> pace and I was able to kill jobs this morning.
> Thanks again.
> Mike
>
> On 8/15/14 9:32 AM, "John Chilton" wrote:
>
>>Very interesting - whenever you get a chance I would try to kill
>>simple jobs when Galaxy is not under l
hg update latest_2014.08.11
To make sure you have the latest and greatest Galaxy and then retry?
If you still have problems and it is the same strack trace - please
let me know. Next up would be to eliminate the proxy theory - not sure
how to do that though :(.
-John
On Wed, Aug 13, 2014 at 4:3
Hello Martin,
You have found something of a bug - there are problems when the output
name matches an input name - in this case your conditional name
"Project" matches your output name "Project" . I think you can just
fix this by calling the output Project_output or the conditional
which_project or
for submit failed: 15007: No permission
>
>
> Iry
>
> On 8/15/14 12:11 PM, "John Chilton" wrote:
>
>>Okay - I just checked out a clean Galaxy without modifications and I
>>don't see any problems with uploads. So this is likely something to do
>>with
new install on one of
>> our other servers and see if the problem occurs there or not.
>>
>> Will update you when we do this.
>>
>> Thanks.
>> Audrey
>>
>>
>> On Fri, Aug 15, 2014 at 8:20 PM, John Chilton wrote:
>>>
>>> Okay - I am
Interesting. I have never seen that behavior before - any chance you
can try this patch and see if it fixes the problem?
https://gist.github.com/jmchilton/b9ae7588de4b21e59cec
-John
On Tue, Aug 19, 2014 at 1:24 PM, Michael Mason
wrote:
> Hi I am running a work flow on data in a history with 54
Are these 17 files fairly dynamic - i.e. are they going to depend on
previous steps in a history or user uploads - or are they literally
just 17 more or less fixed files associated with that tool and every
user should see roughly the same 17 options.
-John
On Thu, Aug 21, 2014 at 3:03 PM, George
dynamic.
> Say all the files of tabular type that the User has at the moment.
>
>
> On Thu, Aug 21, 2014 at 12:28 PM, John Chilton wrote:
>>
>> Are these 17 files fairly dynamic - i.e. are they going to depend on
>> previous steps in a history or user uploads - or
Yes sorry about this - Eric reported this same problem this week.
bioblend is not using API endpoints to download datasets and this is a
problem for proxies with REMOTE_USER configured since most such
proxies are only configured to allow through requests on /api and this
is definitely not the route
Well it looks like the migration file has these columns listed in a
different order than the mapping Galaxy uses - and the order yours
appeared in were the ones from Galaxy's mapping file. So somehow
Galaxy is automatically creating those tables prior to running the
migration based on the code in G
Thanks Enis, just to elaborate on Pulsar - I suspect it would work
with something like configuring Galaxy with S3 object store right now
- but it would do so by having Galaxy cache the data locally and then
Pulsar would negotiate the transfer with Galaxy (many different ways
this could occur depend
Thanks for the bug report. I would say that sounds like it should be
easy to fix - but the last time there was a tool grouping bug - it
took multiple of us hours (days in some cases I think) of
investigation to figure it out - so I have created a Trello card to
track progress on this https://trello
Galaxy's handling of dynamic options has some restrictions for sure
outlined here and it is true that conditional test parameters cannot
be set at runtime in workflow execution (the workflow form doesn't
have the ability to restructure itself based on the conditional the
way the tool form does).
A
This looks like some script is correct locally but not on the cluster.
It looks like the script has some error (like are modules available
locally but not on the cluster). At any rate this error is being
reported on standard error and the script is continuing to run (hence
the datasets being writte
Hello Neil,
The newer uploader targets the tool's API with in browser JavaScript
- so bioblend say targets the same API with its tool client
(https://github.com/afgane/bioblend/blob/master/bioblend/galaxy/tools/__init__.py#L104).
I imagine you could get a similar degree of parallelism by splitti
Collection are one potential answer for how users can specify the set
of stuff that belongs in the directory. For explicitly dealing with
applications that consume directories - I think it is best to just
create the directory and link in files (if possible) before the tool
runs.
mkdir input_dir;
#
There are definitely limitations to the size of workflows that can be
executed right now but I feel like that problem should be getting
better not worse so this is a little confusing. Did something besides
the Galaxy version change (like proxy settings, timeouts, etc...?).
The client side error in
Hello Thomas,
After public debate on Trello, there have been a couple tools that
have indeed disappeared completely but nearly all tools that have gone
away have been migrated to the tool shed and can be recovered with the
migration scripts that are included with each new release
(https://wiki.g
Funny - I just exchanged an e-mail with someone about this tool this
morning, the Galaxy-P dbbuilder tool escapes exactly this santizing.
Here is the relevant portion:
The full tool is on the tool shed.
https://tool
I am not certain order_index solves all of the problems but I agree it
is a step forward and in the latest galaxy distribution the default is
to use 'order_index' instead of the old database id *IF* instead of
specifying inputs via the 'ds_map' parameter you specify the parameter
as 'inputs'. ('inp
Is there some small difference between your test and production
instance? It seems like you client is trying to connect to port 8080
in production? Is this accessible? The error makes me think this is
not a Galaxy problem but a firewall or proxy issue. Is your production
instance served out on port
Hello Peter,
For both of your questions it would be very helpful to know what
version of Galaxy you are using (which tagged version or output of hg
id -i).
I noticed you have min on your repeat - this did not work for the
older Twill test runner historically (and may still not) but it isn't
a
I don't believe this is currently available via the API - but we would
like to move everything to the API so I don't see why this wouldn't be
included (maybe with some rate limit-ting :)). I have created a Trello
card:
https://trello.com/c/W2PEpe8V
This could potentially be fairly straight forwar
I don't think anything like this is currently possible - we are
working on some structural changes to workflow evaluation that should
allow for the addition of stuff like this more easily (right now after
all the jobs are queued up by Galaxy - to some large approximation
Galaxy doesn't really reaso
Hey Evan,
The following diff *should* add a new config option in
universe_wsgi.ini called no_reply_email_from that can be used to set
from when sending these activation emails. If you are feeling brave,
try it out and let me know how it goes. I have not recreated this
problem (read I have not test
Hello David,
We don't have a workaround yet - but we (Dan did most of the
investigating) understand the problem I think. You are using a MySQL
database I assume? (Ignore the rest of this e-mail if not). I think
MySQL has lower limits on the size of BLOBs that can be stored in the
database than pos
I doubt there is a config option somewhere that is going to fix this.
Can your compute nodes download external data outside of Galaxy? (One
could submit a test script to the cluster with a download and see if
it succeeds.) If you give me the URL and let me know what kind of
cluster you are running
I don't know how one would do this precisely but I believe it is
possible at the proxy level - i.e. setting up Apache to handle this
and authenticate with system resources. I believe Tim Booth's work on
integrating galaxy into biolinux uses system users for authentication
by default.
The following
Sorry this has been sitting in my inbox for a while - because I wanted
to try to fix it before the next release but I don't think it is going
to happen.
Generically - it doesn't surprise me that Galaxy can handle certain
parameters changing between versions but not conditionals - they are
handled
I don't know about a standard way of doing - I am not sure about
Galaxy officially supporting this but you can add whatever properties
you want in universe_wsgi.ini (which will be renamed config/galaxy.ini
in coming release). These can be accessed within tool cheetah blocks
as $__app__.config.confi
> As it happens, we also had a few cases with multiple instances of tool-shed
> tools being installed. Since this server is brand new, it was fine to delete
> them. I suppose that maybe we are now fitting in MySQL BLOBs now??
>
> thanks again -
> David
>
>
>
> On
at you would intend to do).
Thanks.
-John
On Wed, Oct 1, 2014 at 1:26 PM, John Chilton wrote:
> I don't know about a standard way of doing - I am not sure about
> Galaxy officially supporting this but you can add whatever properties
> you want in universe_wsgi.ini (which will b
Yes - I am with Iyad - having more context would probably help :). You
can inject job_wrapper related stuff into tool execution in a very
round about way perhaps by creating a dynamic job destination that
consumes the job_wrapper and then creates a job destination with
enviornment variables set to
Thanks for the screenshot.
That page will often and regrettably report errors when there are
none. I think the advice I have been given by people who know the tool
shed better is to click install and then leave that page and go to the
list of installed repositories to track progress.
Did that sam
Hi Marcus,
There was some discussion of proxies a while ago - and it looks like
people were able to get it to work, does this advice help any?
http://dev.list.galaxyproject.org/How-to-install-tool-shed-behind-proxy-with-authentication-td4663218.html
As for the install and test framework - I am p
How would this be different then the third option you listed?
You want it to work with all tools and you as the developer want to be
able to construct these files without needing a dummy tool to produce
the values?
How would imagine these setting files would be disseminated to users
and then sele
I was unable to reproduce this problem. Can you supply more details -
what path did you take through the UI to get that URL? Is that
restricted to just some workflows or do all workflows have this
problem as far as you can tell?
I guess ultimately the reason that workflow isn't downloading is that
? Is that
> not the case anymore ?
>
> Thanks again,
> Wolfgang
>
>
> On 10/13/2014 02:34 AM, John Chilton wrote:
>>
>> I was unable to reproduce this problem. Can you supply more details -
>> what path did you take through the UI to get that URL? Is that
>>
When you say "p.s. there is nothing in the log file."?
Does you mean the Galaxy log file? If there are jobs running and
completing there must be something in the log file right? Do you mean
there are just not tracebacks indicating errors in the logs?
My guess is the response is just taking a whil
I believe there is a fix in the latest update to Galaxy for this thanks to Nate.
https://bitbucket.org/galaxy/galaxy-central/commits/579d211
The previous change to that file was in 2011 - so you are likely to be
able to apply the commit to any release of Galaxy you may be running.
-John
On Tue,
JJ,
Arg this is a mess. I am very sorry about this - I still don't
understand extra_files_path versus files_path myself. There are open
questions on Peter's blast repo and no one ever followed up on my
object store questions about this with Bjoern's issues a couple
release cycles ago. We need to g
thinking about this and respond again.
-John
On Wed, Oct 15, 2014 at 9:39 AM, John Chilton wrote:
> JJ,
>
> Arg this is a mess. I am very sorry about this - I still don't
> understand extra_files_path versus files_path myself. There are open
> questions on Peter's b
r_only=True,
> extra_dir=self._extra_files_path or "dataset_%d_files" % self.id )
>
> I'm trying to see when self._extra_files_path gets set. Otherwise, would
> this return the path relative to the current file location of dataset?
>
>
>
>
> On 10/15/14, 9:36 AM, J
On Wed, Oct 15, 2014 at 1:47 PM, Björn Grüning
wrote:
> Hi John,
>
> glad to see this gets some attention!
>
> Am 15.10.2014 um 19:05 schrieb John Chilton:
>> Hey JJ,
>>
>> Opened a pull request to stable with my best guess at the right to
>> proceed and h
Hmm hopefully someone more knowledgeable about the tool shed than
me responds also but I had a couple quick thoughts.
The first is a warning - workflows may break. At very least workflows
that depend on the previous instance having gone through tool
migrations instead of tool install. My under
In theory GALAXY_SLOTS is now always set unless Galaxy is pretty old -
I still think it is important for tools to specify defaults using the
\${GALAXY_SLOTS:-4} idiom so that tools themselves don't have explicit
dependencies on the Galaxy runtime and in case new resource managers
or job runners com
Hey Pieter,
Sorry I am swamped right now so I don't have time to dig into this
in detail - but I have encountered this before with datatypes that are
compressed - zipped, gzipped, etc Galaxy will attempt to
decompress them in order to figure out what they are. I believe this
is what is happe
Well host is probably not enough - it doesn't seem like a good idea to
assume http right? There is the prefix and (trans.request.scheme) that
could be missing. Is trans.request.base what you want? Maybe
trans.request.path_url? It should be possible to recover that from
trans I guess - though I don'
This feature was added with the August 2014 release and it looks like
you are still on the June release (latest_2014.06.02). You will need
to upgrade to the August or October release (remember to run tool
migrations). (This feature touched a number of different components
and so I don't have a smal
On Fri, Oct 24, 2014 at 10:18 AM, Lukasse, Pieter wrote:
> Hi ,
>
>
>
> I am trying to get a dependency to work in practice (when running a tool
> after the installation). But I keep getting the following message when
> running the tool:
>
>
>
>
>
> [sshexec] galaxy.tools.deps DEBUG 2014-10-24 1
On Wed, Oct 29, 2014 at 7:45 AM, Aleksey Jironkin <
aleksey.jiron...@phe.gov.uk> wrote:
>
> Hi,
>
> I was wondering if there is a way to prevent auto-selecting an option
> for Running a workflow? Quiet often users (especially novices) would
> re-run workflows without checking much the data that goe
Ugh... sorry I missed this e-mail. I don't have any answers yet but I
have created a Trello card to track the issue.
https://trello.com/c/3yVLp8RA
Thanks for the detailed error report.
-John
On Fri, Oct 31, 2014 at 12:16 PM, Peter Cock wrote:
> I have "solved" this by commenting out the appare
I am not aware of any other way new_file_path is accessed. Is it old
jobs that were queued up prior to the switch? I think
job_working_directory might actually be affected by the settings in
your object store configuration. Have you configured an object store?
In either case - knowing the specific
Can you review this dev thread
http://dev.list.galaxyproject.org/Problem-selecting-datasets-with-a-specified-datatype-td4665366.html?
In particular can you go to the "View data types registry" option in
the admin menu and see if Galaxy thinks it knows about this datatype?
Let me know if Galaxy thi
I think a working example that uses this pattern is Bjoern's glimmer tool:
https://github.com/bgruening/galaxytools/tree/master/glimmer3
The only obvious difference I see that he sets the environment
variable to $REPOSITORY_INSTALL_DIR in tool_dependencies.xml
($REPOSITORY_INSTALL_DIR). Does
chan
It looks like this was a problem for awhile but isn't anymore on main
https://trello.com/c/yFeXvhUg.
It doesn't look like there has been any changes to the bowtie2 wrapper
since it was migrated to the tool shed so I am not sure why it isn't
working for you?
As an admin can you open a successful b
Hello Liz,
I wish I had some better news - upgrading might help - we have
definitely made modifications to that file since then - but not to
address this issue I don't think. What would be really interesting is
to see the Galaxy logs around the time of that error - I wonder if
there would be som
I think you want your job_conf.xml destination to have the parameter
true
not
true
Does this help any?
-John
On Thu, Nov 6, 2014 at 1:35 AM, Weiyan Shen wrote:
> Hi,
> I integrated Docker-based tools SMALT whin my local Galaxy by the
> guid:https://github.com/apetkau/galaxy-hackathon-2014/t
Hmm... this is probably a script that works fine on your login node
but not on your worker nodes or vise versa? The fact that it is
writing to standard error and happens for each new shell is probably
what is causing Galaxy jobs to fail - Galaxy is thinking the
underlying applications are writing c
I have no issue with supporting this in general - but the
implementation I think is a bit more tricky than it would seem. The
test framework doesn't know if Galaxy would uncompress zipped files or
not - I think the only way to reason about that in the abstract is to
actually upload the file to Gala
Hello Lance,
I cannot think of a good way to rescue these jobs. If you are
curious about the code where jobs are selected for execution - I would
check out the job handler (lib/galaxy/jobs/handler.py) - see
__monitor_step for instance.
It seems like to prevent this from happening in the futur
Hey Bruno,
So Greg has moved on new exciting things and I am not sure if there
is anyone who now routinely does the tool shed bootstrapping workflow
that can help.
I tried the bootstrapping process and it worked fine with an sqlite
database - maybe it is not compatible with postgres? Want to
Yup - sorry :(. Hopefully by the end of the year I will have something
committed on that - it is a lot of work to do though.
-John
On Wed, Nov 12, 2014 at 4:25 PM, Aaron Petkau wrote:
> I think I may have just found the answer to my question in this trello card
> https://trello.com/c/ndVQmt3G/18
1 - 100 of 486 matches
Mail list logo