Forget about it, recovery got a java.io.IOException: No space left on device but it wasn't clear until i inspected the real logs.
The logs in de web admin didn't show the disk space exception, even when i expand the log line. Maybe that could be changed. Thanks, Markus -----Original message----- > From:Markus Jelsma <markus.jel...@openindex.io> > Sent: Tuesday 24th April 2018 14:39 > To: Solr-user <solr-user@lucene.apache.org> > Subject: IndexFetcher cannot download index file > > Hello, > > Slightly different questions/problem, what is going on here on 7.2.1? During > the recovery, none of this node's fellow replicas indexes were changed but we > still got this error. > > When we got that error, the recovery was restarted, but shortly after the > replicas indexes got updated and recovery failed again. I can understand the > second failure, but not the first. > > Right now the recovering node seems to be stuck, it cannot recover before the > 15 minute indexing cycle is started. Do i really need to stop the indexer and > let the node finish recovery? Can we do something about that? > > Also, why is fetching the full index so slow? It should take just a few > minutes to transfer 50 GB between those nodes. While recovering, CPU > utilization is normal/low. > > Many thanks, > Markus > > Error fetching file,​ doing one > retry...:org.apache.solr.common.SolrException: Unable to download _l5d.cfs > completely. Downloaded 58720256!=218562805 > Error in fetching file: _l5d.cfs (downloaded 58720256 of 218562805 bytes) > Error deleting file: _l5d.cfs > Index fetch failed :org.apache.solr.common.SolrException: Unable to download > _l5d.cfs completely. Downloaded 58720256!=218562805 > Error while trying to recover:org.apache.solr.common.SolrException: > Replication for recovery failed. > Recovery failed - trying again... (0) >
