hey there... for our IRIDA deploys we need to have Singularity images - on Galaxy servers these are either (1) pulled from quay.io using the
mulled_singularity resolver or (2) (for non-IUC tools) built using the build_mulled_singularity resolver (which actually doesn't work for me at the moment). how can this functionality be run on the command line or via API so that I can get the Singularity images for IUC tools (or potentially for non IUC tools) without running things through a Galaxy server?
planemo mull tool_path. If these are to be built for galaxy tools you can register your tools in https://github.com/galaxyproject/planemo-monitor
the build_mulled_singularity resolver works and has tests
1 reply
singularity itself can convert docker images to singularity images, no CLI needed
Can I get a merge on galaxyproject/tools-iuc#4564 ? I think we'll need this to run the tests against 22.05 in the release testing process
Would the iuc welcome wrappers for seqkit (https://bioinf.shenwei.me/seqkit/) tools? I've got a lab that needs some of them, but hoping for feedback since there are several other existing tools that perform similar tasks. Thanks in advance!
Looked at it half a year ago or so, but had no time to do any wrapping work on it. Quite some overlap with tools we have already, but always good to have alternatives :)
More details on that are in https://planemo.readthedocs.io/en/latest/writing_advanced.html#processing-identifiers
Yes that'll work. You probably want to escape some characters as with the regex mentioned in the docs
so...
planemo mull uses involucro to build a Docker contain on the local machine... and singularity build can pull from a Docker registry to build a Singularity image. is there a way to avoid round-tripping through a Docker registry when building Singularity images?
2 replies
@sncgbrl
Hello everyone, I am trying to upload my data but I am getting this #error
(Warning: Error: tus: unexpected response while creating upload, originated from request (method: POST, url: /api/upload/resumable_upload/, response code: 502, response text:
Galaxy
Galaxy could not be reached
You are seeing this message because a request to Galaxy timed out or was refused. This may be a temporary issue which could be resolved by retrying the operation you were performing. If you receive this message repeatedly or for an extended amount of time, please check for additional information on the Galaxy status page or the @galaxyproject Twitter feed. If the issue is not addressed on those sources, you may report it to the support team at galaxy-bugs@galaxyproject.org with details on what you were trying to do and the URL in the address bar.
, request id: n/a)
Can anyone explain to me what is the problem ?
thank you very much in advance.
1 reply
I could produce an interleaved file, subsample with seqtk, then deinterlace ... but that seems more complicated than necessary
I'm having a problem with my Galaxy run in
galaxy serve - I use a local copy of the code and get this error - https://gist.github.com/pvanheus/986a1419d78012448a64a920c0a0df1f - its failing to parse a registry.xml file that is in the job working directory but I don't see where this fie is coming from? (it seems to contain datatypes_conf.xml and more)
2 replies
I'm trying it again without
--galaxy_root to see what happens. waaaaaiting for JS to compile.
I'll have another look if I can figure out how that can happen
pvanheusden: seem similar to what I saw, the registry.xml is supposed to be there but mine was invalid with the actual datatypes configuration copied over many times
creating invalid xml
Tried a couple of things, including changing release, and eventually it went away.
But I really do not know why it happened or why it went away
LOL ok. so I seem to not have it if I don't use
--galaxy_root
It looks a bit like the file was written by multiple writers or so.
yep. and I can't seem to find a cached version of the file anywhere.
mine had a
<sniffer>...</sniffer></datatypes><datatype> in it - so bad XML
mine was invalid xml for sure too, what I meant is that it was a bit too organized for a rogue writer theory
hope all is ok on your sides !
Just see that we have an issue trying to manage a binary file (with datatype not known from Galaxy) in our cheetah code on the XML tool file.... it seems that Galaxy ""don't see"" the file so we can't copy or use it
If we change the input datatype from "binary" to "binary,txt" and that we provide a txt file, "Galaxy see" the file (on the command line generated by Galaxy, we see the file name)
what does mean "not see"? in the UI?
;) sorry for the not precise questoin ;)
in the command line (when with planemo we execute the tool, then look at the job "information" and the "command line"), we have "nothing" where we put the $input (name of the input parameter)
when the file is binary
when the fle is txt, we have for example
input_file.txt
can you point us to some code?
cp $input_raster ${input_raster.element_identifier} &&
cp $input_header ${input_header.element_identifier} &&
Rscript
'$__tool_directory__/data_reading.r'
'$input_raster.element_identifier'
'$input_header.element_identifier'
'$output_shannon'
'$output_alpha'
'$output_beta'
'$plots'
]]>
</command>
<inputs>
<param name="input_raster" type="data" format="binary,txt" label="Input raster"/>
<param name="input_header" type="data" format="txt" label="Input raster header"/>
</inputs>
so here is the code wonder Björn