Regenerate tool files for test.galaxyproject.org

[natefoo]

This will also cause the newest versions of every tool on Test to be installed...

🔥🔥🔥

[natefoo]

Ephemeris is doing a whole lot of lying about what it's installing, which has frequently been my experience installing with Ephemeris in the past. =/

It claimed to install spades successfully:

(8/8) Installing repository spades from nml to section "assembly" at revision b8d633fbf5f5 (TRT: 0:03:17.567326)
	repository spades installed successfully (in 0:00:05.118080) at revision b8d633fbf5f5

But in fact failed miserably:

tool_shed.util.repository_util DEBUG 2019-10-05 05:39:10,967 [p:46,w:1,m:0] [uWSGIWorker1Core2] Adding new row for repository 'spades' in the tool_shed_repository table, status set to 'New'.
tool_shed.galaxy_install.repository_dependencies.repository_dependency_manager DEBUG 2019-10-05 05:39:11,129 [p:46,w:1,m:0] [uWSGIWorker1Core2] Building repository dependency relationships...
galaxy.tools.toolbox.base DEBUG 2019-10-05 05:39:11,129 [p:46,w:1,m:0] [uWSGIWorker1Core2] Appending to tool panel section: Assembly
galaxy.tools.toolbox.base DEBUG 2019-10-05 05:39:11,813 [p:46,w:1,m:0] [uWSGIWorker1Core2] Appending to tool panel section: Assembly
tool_shed.util.hg_util ERROR 2019-10-05 05:39:12,527 [p:46,w:1,m:0] [uWSGIWorker1Core2] Error cloning repository: [Errno 12] Cannot allocate memory
tool_shed.util.repository_util DEBUG 2019-10-05 05:39:12,528 [p:46,w:1,m:0] [uWSGIWorker1Core2] Removed repository installation directory: /cvmfs/test.galaxyproject.org/shed_tools/toolshed.g2.bx.psu.edu/repos/nml/spades/b8d633fbf5f5

Many others failed the same way. Maybe Galaxy is lying to Ephemeris, I don't know.

Also it's not waiting for conda installs to finish once moving from the initial post to status polling:

$ ps uxw | grep conda
g2test    4820 87.7 15.9 741172 620948 ?       R    01:34   2:33 /cvmfs/test.galaxyproject.org/deps/_conda/bin/python /cvmfs/test.galaxyproject.org/deps/_conda/bin/conda create -y --quiet --override-channels --channel iuc --channel conda-forge --channel bioconda --channel defaults --name mulled-v1-402b40ac33a113bff7165f013c30adb879fd15a97f823dde0b9e60322b8bb7d5 augustus=3.2.3 maker=2.31.10
g2test    4872 41.5 14.0 665772 545168 ?       S    01:36   0:40 /cvmfs/test.galaxyproject.org/deps/_conda/bin/python /cvmfs/test.galaxyproject.org/deps/_conda/bin/conda create -y --quiet --override-channels --channel iuc --channel conda-forge --channel bioconda --channel defaults --name __unicycler@0.4.7 unicycler=0.4.7
g2test    4909  3.6  0.4 104040 19056 ?        S    01:37   0:00 /cvmfs/test.galaxyproject.org/deps/_conda/bin/python /cvmfs/test.galaxyproject.org/deps/_conda/bin/conda create -y --quiet --override-channels --channel iuc --channel conda-forge --channel bioconda --channel defaults --name __quast@5.0.2 quast=5.0.2

Which is probably the cause of all the memory errors.

[natefoo]

Pretty sure this here is gonna match any version already installed.

[natefoo]

Fix in galaxyproject/ephemeris#146.

[martenson]

@natefoo https://github.com/galaxyproject/ephemeris/blob/master/src/ephemeris/shed_tools.py#L462

[natefoo]

😭

[natefoo]

@galaxybot deploy this

[martenson]

fyi the log of one build of this on jenkins is 88 MB, we need to keep proper cleaning in mind

[natefoo]

Not sure this method we're using is ideal. The recursive ls is hard to scan. Although these are some of the largest sets of changes we are likely to ever have.