So you are wrestling with aliases. You can not delete aliases by file
system operations. Have you checked

http://www.elasticsearch.org/guide/en/elasticsearch/reference/current/indices-aliases.html#deleting

for deleting aliases?

Jörg


On Tue, Aug 12, 2014 at 4:10 AM, Sam2014 <[email protected]> wrote:

> ///Attached is the cluster log that covers the whole sequence from
> shutting the cluster and re-activating it.
>
> //Show me current state of cluster
> {
>   "cluster_name": "rexCluster",
>   "status": "red",
>   "timed_out": false,
>   "number_of_nodes": 2,
>   "number_of_data_nodes": 2,
>   "active_primary_shards": 7,
>   "active_shards": 7,
>   "relocating_shards": 0,
>   "initializing_shards": 2,
>   "unassigned_shards": 21
> }
>
> //Show me indices on the cluster
>
> GET _aliases
> {
>    "foo": {
>       "aliases": {}
>    },
>    "mediregs": {
>       "aliases": {}
>    },
>    "twitter": {
>       "aliases": {}
>    }
> }
>
>
>
> //Shut down the cluster
>
> POST /_cluster/nodes/_shutdown
> {
>    "cluster_name": "rexCluster",
>    "nodes": {
>       "_vPEfmc9TruIpugk4NWleg": {
>          "name": "node 1"
>       },
>       "DB94rWiiSvS9HUTXwI7GTw": {
>          "name": "node 2"
>       }
>    }
> }
>
> //Go into the fs and delete files under /data
>
> ubuntu@domU-12-31-39-0A-99-EA:/opt/cores/elasticsearch/data/rexCluster/nodes/0$
> ls
>
> *indices*  node.lock  *_state*
>
> ubuntu@domU-12-31-39-0A-99-EA:/opt/cores/elasticsearch/data/rexCluster/nodes/0$
> cd indices/
>
> ubuntu@domU-12-31-39-0A-99-EA:/opt/cores/elasticsearch/data/rexCluster/nodes/0/indices$
> ls
>
> *foo*  *mediregs*  *twitter*
>
> ubuntu@domU-12-31-39-0A-99-EA:/opt/cores/elasticsearch/data/rexCluster/nodes/0/indices$
> sudo rm -rf foo
>
> ubuntu@domU-12-31-39-0A-99-EA:/opt/cores/elasticsearch/data/rexCluster/nodes/0/indices$
> sudo rm -rf mediregs
>
> ubuntu@domU-12-31-39-0A-99-EA:/opt/cores/elasticsearch/data/rexCluster/nodes/0/indices$
> sudo rm -rf twitter
>
> ubuntu@domU-12-31-39-0A-99-EA:/opt/cores/elasticsearch/data/rexCluster/nodes/0/indices$
> ls
>
> ubuntu@domU-12-31-39-0A-99-EA
> :/opt/cores/elasticsearch/data/rexCluster/nodes/0/indices$
> //Restart nodes 1 &2 - should be in log file
>
> Get _aliases //Again - they came back!!
> {
>    "foo": {
>       "aliases": {}
>    },
>    "mediregs": {
>       "aliases": {}
>    },
>    "twitter": {
>       "aliases": {}
>    }
> }
>
>
> //Ok - Try to delete index
> DELETE /mediregs
>
> //takes little while and comes back
> {
>    "acknowledged": false
> }
>
>
>
>
> On Monday, August 11, 2014 7:12:22 PM UTC-4, Jörg Prante wrote:
>
>> Can you post your detailed steps and debug logs?
>>
>> Make also sure there is no third node - this looks like ES is shipping
>> shards to a third node from which the two nodes are recovering from.
>>
>> Jörg
>>
>>
>> On Tue, Aug 12, 2014 at 1:06 AM, Sam2014 <[email protected]> wrote:
>>
>>> I stopped node 1, then I stopped node 2. I disabled the cluster, deleted
>>> files from nodes, Made sure no Indices on either node. I re-enabled cluster
>>> in yml, restarted cluster, and the index came back!
>>>
>>>
>>> On Monday, August 11, 2014 6:16:42 PM UTC-4, Jörg Prante wrote:
>>>
>>>> There is no vicious circle. First you must stop the cluster. Then,
>>>> after processes have exited, continue with file operations. After that, you
>>>> can start the cluster nodes again.
>>>>
>>>> Jörg
>>>>
>>>>
>>>> On Mon, Aug 11, 2014 at 11:49 PM, Sam2014 <[email protected]> wrote:
>>>>
>>>>> Oops, When I restarted the cluster/nodes, I see shards under an
>>>>> "unassigned node", neither of the two nodes of the cluster. Is there a way
>>>>> I can route/assign /delete out of this vicious cycle?
>>>>>
>>>>>
>>>>> On Monday, August 11, 2014 5:46:25 PM UTC-4, Sam2014 wrote:
>>>>>>
>>>>>> I also deleted the directory, restarted the nodes, and it recovered
>>>>>> the index still, which is null.
>>>>>>
>>>>>> On Monday, August 11, 2014 5:42:53 PM UTC-4, Sam2014 wrote:
>>>>>>>
>>>>>>> I wish it was that simple, I deleted the directory of the index from
>>>>>>> /data. Yet, the index is still there and won't be deleted or 
>>>>>>> recreated.. I
>>>>>>> still get acknowledged = false.
>>>>>>>
>>>>>>> On Monday, August 11, 2014 5:31:36 PM UTC-4, Jörg Prante wrote:
>>>>>>>>
>>>>>>>> Your JVMs (at least one of them) are not able to process the
>>>>>>>> command, so you should check out the data directory where the file of 
>>>>>>>> the
>>>>>>>> index to be deleted are located. Shut your cluster down, remove the 
>>>>>>>> index
>>>>>>>> files of the index to be deleted from the filesystem, and start the 
>>>>>>>> cluster
>>>>>>>> up again.
>>>>>>>>
>>>>>>>> To avoid such situations, find out why the JVMs could not react,
>>>>>>>> most probably there are exceptions in the server logs, and fix the 
>>>>>>>> trouble.
>>>>>>>>
>>>>>>>> Jörg
>>>>>>>>
>>>>>>>>
>>>>>>>> On Mon, Aug 11, 2014 at 11:24 PM, Sam2014 <[email protected]>
>>>>>>>> wrote:
>>>>>>>>
>>>>>>>>> I got myself in trouble by changing the mapping.
>>>>>>>>>
>>>>>>>>> I am on a cluster of 2 nodes, and trying to delete the index, I
>>>>>>>>> keep getting acknowleged = false with normal delete curl -X GET "
>>>>>>>>> http://localhost:9200/myIndex
>>>>>>>>>
>>>>>>>>> If I look at my index through HEAD plugin, I get
>>>>>>>>> "myIndex"
>>>>>>>>> size: unknown
>>>>>>>>> docs: unknown
>>>>>>>>>
>>>>>>>>> and I see no shards assigned to either node apparently (Everything
>>>>>>>>> in this index indicates Null). I would just like to be able to delete 
>>>>>>>>> it.
>>>>>>>>> How can I force a delete?
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>  --
>>>>>>>>> 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 [email protected].
>>>>>>>>> To view this discussion on the web visit
>>>>>>>>> https://groups.google.com/d/msgid/elasticsearch/c6818593-550
>>>>>>>>> 0-4762-9f59-46a2e2a3f772%40googlegroups.com
>>>>>>>>> <https://groups.google.com/d/msgid/elasticsearch/c6818593-5500-4762-9f59-46a2e2a3f772%40googlegroups.com?utm_medium=email&utm_source=footer>
>>>>>>>>> .
>>>>>>>>> For more options, visit https://groups.google.com/d/optout.
>>>>>>>>>
>>>>>>>>
>>>>>>>>  --
>>>>> 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 [email protected].
>>>>> To view this discussion on the web visit https://groups.google.com/d/
>>>>> msgid/elasticsearch/8e616783-23fe-4029-9a53-6e30389abe6c%40goo
>>>>> glegroups.com
>>>>> <https://groups.google.com/d/msgid/elasticsearch/8e616783-23fe-4029-9a53-6e30389abe6c%40googlegroups.com?utm_medium=email&utm_source=footer>
>>>>> .
>>>>>
>>>>> For more options, visit https://groups.google.com/d/optout.
>>>>>
>>>>
>>>>  --
>>> 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 [email protected].
>>> To view this discussion on the web visit https://groups.google.com/d/
>>> msgid/elasticsearch/255a872f-02d3-49e5-917e-9736ffcf5c3b%
>>> 40googlegroups.com
>>> <https://groups.google.com/d/msgid/elasticsearch/255a872f-02d3-49e5-917e-9736ffcf5c3b%40googlegroups.com?utm_medium=email&utm_source=footer>
>>> .
>>>
>>> For more options, visit https://groups.google.com/d/optout.
>>>
>>
>>  --
> 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 [email protected].
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/elasticsearch/980ceb4c-1312-4844-8d70-00f85d39b51e%40googlegroups.com
> <https://groups.google.com/d/msgid/elasticsearch/980ceb4c-1312-4844-8d70-00f85d39b51e%40googlegroups.com?utm_medium=email&utm_source=footer>
> .
>
> For more options, visit https://groups.google.com/d/optout.
>

-- 
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 [email protected].
To view this discussion on the web visit 
https://groups.google.com/d/msgid/elasticsearch/CAKdsXoG6e%3D9jUxtXxkV1xopLjPqwYxhzZ76L3URRL%2BEhznPOKQ%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.

Reply via email to