As far as I remember you can only delete slices which are either 1. Inactive or 2. A part of a collection that uses implicit routing.
The delete shard API would not let you delete a slice directly. However, you can perhaps retry the command. As far as I remember, it's built to cleanup an earlier incomplete run on subsequent calls. As for a requestParam that allows you to specify a higher timeout value, I don't think that's in place yet. You can follow the SOLR Jira I mentioned in my last response to track the support of async calls in Solr which should fix this kind of issues. On Fri, Jan 3, 2014 at 4:27 PM, sriram <srira...@ask.com> wrote: > Thanks a lot Anshum for your quick reply. I checked the clusterstate.json > and > there it says the two new shards as under construction. > > It shows : > shard1 : active > shard2 : active > shard1_1 : construction > shard1_0 : construction > > But the loggers don't indicate any other progress in that. Also when I try > to delete the newly created shards, it is not able to delete ( which I > think > should be the case as it is still under construction) > > Is there any workaround to resolve this for now? like some reqestParam > specifying a higher timeout value? > > > > -- > View this message in context: > http://lucene.472066.n3.nabble.com/SPLITSHARD-failed-to-invoke-SPLIT-core-admin-command-tp4109431p4109441.html > Sent from the Solr - User mailing list archive at Nabble.com. > -- Anshum Gupta http://www.anshumgupta.net
