cassandra-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Roland Otta <Roland.O...@willhaben.at>
Subject Re: force processing of pending hinted handoffs
Date Thu, 13 Apr 2017 07:40:16 GMT
i figured out that there is an mbean
org.apache.cassandra.db.type=HintedHandoffManager with the operation
scheduleHintDelivery

i guess thats what i would need in that case. at least the docs let me
think so http://javadox.com/org.apache.cassandra/cassandra-all/3.0.0/or
g/apache/cassandra/db/HintedHandOffManagerMBean.html

but everytime i try invoking that operation i get an
UnsupportedOperationException (tried it with hostname, ip and host-id
as parameters - everytime the same exception)



On Tue, 2017-04-11 at 07:40 +0000, Roland Otta wrote:
> hi,
> 
> sometimes we have the problem that we have hinted handoffs (for
> example
> because auf network problems between 2 DCs) that do not get processed
> even if the connection problem between the dcs recovers. Some of the
> files stay in the hints directory until we restart the node that
> contains the hints.
> 
> after the restart of cassandra we can see the proper messages for the
> hints handling
> 
> Apr 11 09:28:56 bigd006 cassandra: INFO  07:28:56 Deleted hint file
> c429ad19-ee9f-4b5a-abcd-1da1516d1003-1491895717182-1.hints
> Apr 11 09:28:56 bigd006 cassandra: INFO  07:28:56 Finished hinted
> handoff of file c429ad19-ee9f-4b5a-abcd-1da1516d1003-1491895717182-
> 1.hints to endpoint c429ad19-ee9f-4b5a-abcd-1da1516d1003
> 
> is there a way (for example via jmx) to force a node to process
> outstanding hints instead of restarting the node?
> does anyone know whats the cause for not retrying to process those
> hints automatically?
> 
> br,
> roland
> 
> 
Mime
View raw message