Chef-solo monit and elasticsearch version issues

Jan_Backes · March 22, 2014, 12:08pm

Hi,

I am currently trying to install elasticsearch on amazon (amazon ami) using
the chef-solo tutorial (
http://www.elasticsearch.org/tutorials/deploying-elasticsearch-with-chef-solo/).
I followed this tutorial couple months ago and was successfully running an
ES cluster since then.

While doing a "test-run" I realised that the monit version is rather old
and that it is not inline any longer with the updated cookbook resulting in
an error (e.g.
https://github.com/elasticsearch/cookbook-elasticsearch/issues/184). I
changed this by pulling the newest version from
https://github.com/apsoto/monit and it installs just fine.

Now I am just wondering why I am getting a "green" for the cluster health
check:

Remote Host 'elasticsearch_cluster_health'
status online with all services
monitoring status monitored
port response time 0.008s to
0.0.0.0:9200/_cluster/health[HTTP via TCP]
data collected Sat Mar 22 11:39:32 2014

Should the cluster health check fail since it is only one node?

I also tried to update the version to 1.0.1. For that I spun up another
instance using the same settings - but changing the version number in

github.com

elastic/cookbook-elasticsearch/blob/master/attributes/default.rb

# empty settings (populate these for the elasticsearch::default recipe)
# see the resources or README.md to see what you can pass here.
default['elasticsearch']['user'] = {}
default['elasticsearch']['install'] = {}
default['elasticsearch']['configure'] = {}
default['elasticsearch']['service'] = {}
default['elasticsearch']['plugin'] = {}

# platform_family keyed download URLs
default['elasticsearch']['download_urls'] = {
  'debian' => 'https://artifacts.elastic.co/downloads/elasticsearch/elasticsearch-%s.deb',
  'rhel' => 'https://artifacts.elastic.co/downloads/elasticsearch/elasticsearch-%s.rpm',
  'tarball' => 'https://artifacts.elastic.co/downloads/elasticsearch/elasticsearch-%s.tar.gz',
}

# platform_family keyed download sha256 checksums
default['elasticsearch']['checksums']['5.0.0']['debian'] = '0a4f8842a1f7d7bd3015118298284383efcd4c25f9591c46bb5ab68189815268'
default['elasticsearch']['checksums']['5.0.0']['rhel'] = 'fb502a9754f2162f27590125aecc6b68fa999738051d8f230c4da4ed6deb8d62'
default['elasticsearch']['checksums']['5.0.0']['tarball'] = 'a866534f0fa7428e980c985d712024feef1dee04709add6e360fc7b73bb1e7ae'

This file has been truncated. show original

1.0.1. I saw that there are many issues/pull request regarding the
upgrade - so I was thinking that a fresh install would avoid all this.

The bootstrap script and everything runs fine - but the es service will not
start. Trying to start es manually show the error:

{1.0.1}: Initialization Failed ...

ElasticsearchIllegalStateException[Failed to obtain node lock, is the
following location writable?:
[/usr/local/elasticsearch-1.0.1/data/elasticsearch]]
IOException[failed to obtain lock on
/usr/local/elasticsearch-1.0.1/data/elasticsearch/nodes/49]
IOException[Cannot create directory:
/usr/local/elasticsearch-1.0.1/data/elasticsearch/nodes/49]

Now I am wondering how to adjust the cookbook to set the paths right and
avoid the lock error. Or is there another preferred way of upgrading?

I also tried this using ubuntu server (using the updated setup instruction
etc from the elasticsearch-cookbook github page) and ran into the same
issues. Maybe somebody has similar issues/experiences.

Thanks a lot for your help,
Jan

--
You received this message because you are subscribed to the Google Groups "elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send an email to elasticsearch+unsubscribe@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/elasticsearch/CAKzGBByCxf0ny9MYQOqMCV%3DT5RofSz6ziwwKLd5CsC4AWcDbTQ%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.