ActionRequestValidationException on deleteIndex() during beforeEach()

[martinb3]

Hello --

I'm noticing the following in IndexCrud.scala:

trait DeleteIndex {
  self: Indexer =>
  def deleteIndex(indices: Iterable[String] = Nil, timeout: Option[String] = None) = deleteIndex_send(indices, timeout).actionGet
  def deleteIndex_send(indices: Iterable[String] = Nil, timeout: Option[String] = None) = deleteIndex_prepare(indices, timeout).execute
  def deleteIndex_prepare(indices: Iterable[String] = Nil, timeout: Option[String] = None) = {
    val request = client.admin.indices.prepareDelete(indices.toArray: _*)
    timeout foreach { request.setTimeout(_) }
    request
  }

  /** caveat: if none of the indices exists, we will encounter a validation exception */
  def deleteIndexIfExists(indices: Iterable[String] = Nil, timeout: Option[String] = None) = deleteIndexIfExists_send(indices, timeout).actionGet
  def deleteIndexIfExists_send(indices: Iterable[String] = Nil, timeout: Option[String] = None) = deleteIndexIfExists_prepare(indices, timeout).execute
  def deleteIndexIfExists_prepare(indices: Iterable[String] = Nil, timeout: Option[String] = None) = {
    val existing = indices filter (exists(_).isExists)
    val toDelete: Iterable[String] = if (existing.isEmpty) null else existing
    deleteIndex_prepare(toDelete, timeout)
  }
}

How are these tests ever passing if they throw an exception on delete() when I'm starting from scratch with no indexes? Are you doing some special prep step that I'm not catching?

Here's the exception:

org.elasticsearch.action.ActionRequestValidationException: Validation Failed: 1: index / indices is missing;

Thank you,

Martin

[martinb3]

Just a note -- I do think some of this is related to bumping the version of Elasticsearch. However, even going back to your upstream version, I get a lot of failures like:

[info] SpecificMasterNodesTests:
[info] - simpleOnlyMasterNodeElection *** FAILED ***
[info]   org.elasticsearch.discovery.MasterNotDiscoveredException: waited for [30s]
[info]   at org.elasticsearch.action.support.master.TransportMasterNodeOperationAction$3.onTimeout(TransportMasterNodeOperationAction.java:180)
[info]   at org.elasticsearch.cluster.service.InternalClusterService$NotifyTimeout.run(InternalClusterService.java:483)
[info]   at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1145)
[info]   at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:615)
[info]   at java.lang.Thread.run(Thread.java:744)
[info]   ...

[info] FullRollingRestartTests:
[info] - testFullRollingRestart *** FAILED ***
[info]   true did not equal false (FullRollingRestartTests.scala:15)
[info] IndicesStoreTests:
[info] - shardsCleanup *** FAILED ***
[info]   false did not equal true (IndicesStoreTests.scala:23)

etc.

[martinb3]

Hello! I found just adding some delays at specific tests that are doing recovery caused all tests to pass. These commits make all tests pass for me. Thank you! - Martin

[bsadeh]

yes, it might be cpu dependent. i'm pretty sure it didn't fail for me
before on my laptop.
glad you figured out a fix :-)

On Mon, Apr 21, 2014 at 8:09 PM, Martin Smith notifications@github.comwrote:

Hello! I found just adding some delays at specific tests that are doing
recovery caused all tests to pass. These commits make all tests pass for
me. Thank you! - Martin

—
Reply to this email directly or view it on GitHubhttps://github.com//pull/40#issuecomment-40953270
.

[martinb3]

Would you mind merging these changes in? Thanks! - Martin

[bsadeh]

I did merge the last pull request Martin, but I can't release while I
travel.
3 more weeks ...

On Sat, Apr 26, 2014 at 10:26 PM, Martin Smith notifications@github.comwrote:

Would you mind merging these changes in? Thanks! - Martin

—
Reply to this email directly or view it on GitHubhttps://github.com//pull/40#issuecomment-41478222
.

[martinb3]

Thanks! No problem about the release, as I'm only running tests from source :)