[AMBARI-8468] Value of fs.defaultFS predefined in GlusterFS stack doesn't make sense from GlusterFS perspective

• Type: Bug
• Status: Open
• Priority: Minor
• Resolution: Unresolved
• Affects Version/s: 1.6.1
• Component/s: ambari-server, stacks
• Labels: glusterfs hcfs

Environment

HDP 2.1 on RHEL 6 with 2.1.GlusterFS stack.

• Assignee: Unassigned
• Reporter: Martin Bukatovic
• Created: 27/Nov/14 16:19
• Updated: 11/Dec/14 17:13

Description

Default value of fs.defaultFS property as defined in core-site.xml of
2.1.GlusterFS stack is not valid from GlusterFS perspective.

from 2.1.GlusterFS/services/GLUSTERFS/configuration/core-site.xml

  <property>
    <name>fs.defaultFS</name>
    <value>glusterfs:///localhost:8020</value>
  </property>

Leaving the current default value there may create problems for some use cases.
Eg. Hive has a problem with that. Or see for example traceback from Ambari
itself:

from http://ambari-server.example.com:8080/#/main/services/GLUSTERFS/summary

Error: E0904 : E0904: Scheme [glusterfs] not supported in uri [glusterfs:///localhost:8020/user/ambari-qa/examples/apps/map-reduce]
Invalid sub-command: Missing argument for option: info

use 'help [sub-command]' for help details
Invalid sub-command: Missing argument for option: info

use 'help [sub-command]' for help details

workflow_status=
2014-11-14 14:11:51,400 - Error while executing command 'service_check':
Traceback (most recent call last):
  File "/usr/lib/python2.6/site-packages/resource_management/libraries/script/script.py", line 111, in execute
    method(env)
  File "/var/lib/ambari-agent/cache/stacks/HDP/2.0.6/services/OOZIE/package/scripts/service_check.py", line 31, in service_check
    oozie_smoke_shell_file( smoke_test_file_name)
  File "/var/lib/ambari-agent/cache/stacks/HDP/2.0.6/services/OOZIE/package/scripts/service_check.py", line 54, in oozie_smoke_shell_file
    logoutput = True
  File "/usr/lib/python2.6/site-packages/resource_management/core/base.py", line 148, in __init__
    self.env.run()
  File "/usr/lib/python2.6/site-packages/resource_management/core/environment.py", line 149, in run
    self.run_action(resource, action)
  File "/usr/lib/python2.6/site-packages/resource_management/core/environment.py", line 115, in run_action
    provider_action()
  File "/usr/lib/python2.6/site-packages/resource_management/core/providers/system.py", line 239, in action_run
    raise ex
Fail: Execution of '/tmp/oozieSmoke2.sh redhat /etc/oozie/conf /etc/hadoop/conf ambari-qa False' returned 1. 14/11/14 14:11:23 INFO glusterfs.GlusterVolume: Initializing gluster volume..

The obvious fix of this would be to remove the hostname and port from the
default value, so that it would be just glusterfs:///.

The problem is that this can't be done without consequences: some pre-install
scripts would have a problem with that (because we share/reuse code with HDFS
there).

Current workaround is to manually change the value after installation, which is
hardly suitable solution from the long term perspective.

To sum it up: in current Ambari, we can't ship valid default configuration
in GlusterFS stack without breaking pre-install scripts at the same time.

Since this may be usefull for any HCFS (Hadoop Compatible File
System), not just GlusterFS, I propose to change the code in ambari so that
any HCFS can have valid configuration in the configuration templates without
breaking given HCFS or Ambari scripts itself. This would make possible to have
reasonable value of fs.defaultFS in GlusterFS stack.

Issue Links

• relates to: AMBARI-8658 GlusterFS stack doesn't allow to change fs.defaultFS