We ran into this issue as well See the this ticket https://issues.apache.org/jira/browse/PHOENIX-6187 The issue is the OperationStatus returned by region server gets ignore here https://github.com/apache/phoenix/blob/master/phoenix-core/src/main/java/org/apache/phoenix/hbase/index/Indexer.java#L314 I’ll post a patch for it soon
On 2020/10/10 18:12:54, Jins George <[email protected]> wrote: > Hello Community,> > > > > We are facing a wired issue in production with Phoenix write operations. We > use phoenix query server to UPSERT/DELETE records and all the calls returns > success. However intermittently some records are missing in Phoenix/Hbase. > There is no error or exception reported in phoenix query server logs or > Hbase Region server logs, no abnormality found in Flush operations also. We > have auto commit enabled in the client side.> > > > > Any insights into what may be happening is highly appreciated. We had > enable HDFS Name node HA automatic Failover recently, not sure if that has > any impact. This is a cluster running for more than 1 year and never faced > this issue before.> > > > > Phoenix Version: APACHE_PHOENIX-4.14.0, HBase version: 1.2 on CDH 5.11.2> > > > > Thanks,> > > Jins George> > > Data Engineer,> > > Aeris Communications> > >
