nsoranzo

#!/usr/bin/env python

import argparse
import os
import shlex
import subprocess
import sys

scripts_dir = os.path.dirname(__file__)
sys.path.insert(1, os.path.join(scripts_dir, os.pardir, 'lib'))

nsoranzo

$ sudo yum install -y GeoIP-devel gcc gd-devel gperftools-devel libxslt-devel openssl-devel pcre-devel perl-devel perl-ExtUtils-Embed rpm-build zlib-devel
$ mkdir -p ~/software
$ cd ~/software/
[software]$ git clone --branch 2.2 https://github.com/vkholodkov/nginx-upload-module
# Download latest source RPM from EPEL repository https://apps.fedoraproject.org/packages/nginx/
[software]$ wget https://dl.fedoraproject.org/pub/epel/7/SRPMS/n/nginx-1.10.2-1.el7.src.rpm
[software]$ rm -rf ~/rpmbuild/
[software]$ mkdir ~/rpmbuild
[software]$ rpm -i nginx-1.10.2-1.el7.src.rpm
[software]$ cd ~/rpmbuild/

nsoranzo

If a handler log is full of messages like:

Workflow step 1585 of invocation 554 delayed

then probably a workflow invocation is stuck because a step was cancelled by a user. After verifying that this is the case, to stop the workflow invocation scheduling connect to the Galaxy database using psql and do the following:

galaxy_db=> SELECT * FROM workflow_invocation WHERE id=554;
 id  |        create_time         |        update_time         | workflow_id | history_id | state | scheduler | handler  |               uuid               
-----+----------------------------+----------------------------+-------------+------------+-------+-----------+----------+----------------------------------
 554 | 2016-11-16 17:24:38.586902 | 2016-11-16 17:26:34.555872 |         201 |        485 | ready | core      | handler1 | 8d49678eac2111e697b90050569af9e3