[SLIDER-337] Agent error when trying to deploy Accumulo

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Fix Version/s: Slider 0.60
• Component/s: agent-provider

Environment

Gentoo Linux, develop branch (035ebbd), Hadoop-2.6.0-SNAPSHOT (r1619002)

• Assignee: Josh Elser
• Reporter: Josh Elser
• Created: 20/Aug/14 03:59
• Updated: 30/Oct/14 22:03
• Resolved: 20/Aug/14 23:33

Description

Tried to start an Accumulo cluster:

slider create accumulo --image hdfs://localhost:8020/slider/agent/slider-agent.tar.gz --template /home/elserj/slider/app-packages/accumulo/target/apache-slider-accumulo-1.6.0-app-package-0.40/appConfig.json --resources /home/elserj/slider/app-packages/accumulo/target/apache-slider-accumulo-1.6.0-app-package-0.40/resources.json

SliderAM came up, but no Accumulo procs got started. Looking at the AM webapp, the Accumulo master container kept cycling. Dug around in the NM/container logs dir, and found the following stack:

Traceback (most recent call last):
  File ".../nm-local-dir/usercache/elserj/appcache/application_1408503547578_0002/container_1408503547578_0002_01_000021/app/definition/package/scripts/accumulo_master.py", line 24, in <module>
    AccumuloScript('master').execute()
  File ".../nm-local-dir/usercache/elserj/appcache/application_1408503547578_0002/container_1408503547578_0002_01_000021/infra/agent/slider-agent/resource_management/libraries/script/script.py", line 114, in execute
    method(env)
  File ".../nm-local-dir/usercache/elserj/appcache/application_1408503547578_0002/filecache/19/apache-slider-accumulo-1.6.0-app-package-0.40.zip/package/scripts/accumulo_script.py", line 82, in start
    self.configure(env) # for security
  File ".../nm-local-dir/usercache/elserj/appcache/application_1408503547578_0002/filecache/19/apache-slider-accumulo-1.6.0-app-package-0.40.zip/package/scripts/accumulo_script.py", line 76, in configure
    setup_conf_dir(name=self.component)
  File ".../tmp/nm-local-dir/usercache/elserj/appcache/application_1408503547578_0002/filecache/19/apache-slider-accumulo-1.6.0-app-package-0.40.zip/package/scripts/accumulo_configuration.py", line 31, in setup_conf_dir
    recursive = True
  File ".../nm-local-dir/usercache/elserj/appcache/application_1408503547578_0002/container_1408503547578_0002_01_000021/infra/agent/slider-agent/resource_management/core/base.py", line 148, in __init__
    self.env.run()
  File ".../nm-local-dir/usercache/elserj/appcache/application_1408503547578_0002/container_1408503547578_0002_01_000021/infra/agent/slider-agent/resource_management/core/environment.py", line 149, in run
    self.run_action(resource, action)
  File ".../nm-local-dir/usercache/elserj/appcache/application_1408503547578_0002/container_1408503547578_0002_01_000021/infra/agent/slider-agent/resource_management/core/environment.py", line 109, in run_action
    resource.provider)
  File ".../nm-local-dir/usercache/elserj/appcache/application_1408503547578_0002/container_1408503547578_0002_01_000021/infra/agent/slider-agent/resource_management/core/providers/__init__.py", line 72, in find_provider
    if resource in provider[env.system.os_family]:
KeyError: 'gentoo base system'

I'm rusty on my Python, but I believe this means that in slider-agent/src/main/python/resource_management/core/providers/_init_.py

if resource in provider[env.system.os_family]:

should be

if env.system.os_family in provider:
  if resource in provider[env.system.os_family]:

to avoid the Exception when the key doesn't exist (alternatively try/except, too), and get down to the default case.