update_metrics.py: "IndexAlreadyExistsException"

[andral]

Hi

I have elasticsearch running on the same host as graph-exlporer.

When I run update-metrics.py for the first time it creates a new es index but then exceptions out with the following:

2014-01-13 15:51:46,522 - update_metrics - INFO - fetching/saving metrics from graphite...
2014-01-13 15:51:46,543 - update_metrics - INFO - generating structured metrics data...
2014-01-13 15:51:46,543 - update_metrics - DEBUG - loading metrics
2014-01-13 15:51:46,547 - update_metrics - DEBUG - removing outdated targets
2014-01-13 15:51:46,547 - update_metrics - DEBUG - making sure index exists..
2014-01-13 15:51:47,222 - update_metrics - DEBUG - making sure shard is started..
2014-01-13 15:51:47,238 - update_metrics - DEBUG - shard[0][0] state: RECOVERING
2014-01-13 15:51:47,344 - update_metrics - DEBUG - shard[0][0] state: STARTED
2014-01-13 15:51:47,345 - update_metrics - DEBUG - making sure index exists..
2014-01-13 15:51:47,390 - update_metrics - ERROR - sorry, something went wrong: (ElasticException(...), 'ElasticSearch Error: {"error":"IndexAlreadyExistsException[[graphite_metrics] already exists]","status":400}')
Traceback (most recent call last):
  File "./update_metrics.py", line 36, in <module>
    backend.update_data(s_metrics)
  File "/opt/graph-explorer/backend.py", line 78, in update_data
    s_metrics.remove_metrics_not_in(metrics)
  File "/opt/graph-explorer/structured_metrics/__init__.py", line 201, in remove_metrics_not_in
    for es_metrics in self.get_all_metrics():
  File "/opt/graph-explorer/structured_metrics/__init__.py", line 298, in get_all_metrics
    self.assure_index()
  File "/opt/graph-explorer/structured_metrics/__init__.py", line 180, in assure_index
    self.es.post('graphite_metrics', data=body)
  File "/usr/lib/python2.6/site-packages/rawes/elastic.py", line 62, in post
    return self.request('post', path, **kwargs)
  File "/usr/lib/python2.6/site-packages/rawes/elastic.py", line 84, in request
    return self.connection.request(method, new_path, **kwargs)
  File "/usr/lib/python2.6/site-packages/rawes/http_connection.py", line 48, in request
    return self._decode(response, json_decoder)
  File "/usr/lib/python2.6/site-packages/rawes/http_connection.py", line 61, in _decode
    result=decoded, status_code=response.status_code)
ElasticException: (ElasticException(...), 'ElasticSearch Error: {"error":"IndexAlreadyExistsException[[graphite_metrics] already exists]","status":400}')

ES is fine according to http://.../_status. If I run the script again, the same exception gets thrown, this is also true if I delete the index first and then run the script again.

This is on RHEL6.4 with python 2.6.6 and elasticsearch 0.90.10.

Any help is appreciated! Cheers

[Dieterbe]

hmmm so that has been fixed in git for a while, but i haven't done a release recently. i just realised most people will try to use the most recent stable release and hence hit this bug :l

for now you can pretty safely use the latest git version but i guess i really need to do a new release

[Dieterbe]

hey @andral I just pushed a new release that comes with the fix https://github.com/vimeo/graph-explorer/releases/tag/v1.2.0

please reopen if you get the issue again (very unlikely :)

[andral]

Hey @Dieterbe, thanks a lot for the new release, the issue is gone and the script runs without any errors, yay :)

Thanks a lot