I figured out the problem. We have phoenix.upsert.batch.size set to 10 in hbase-site.xml but somehow that property is **not getting picked up in our oozie workflow** When i am explicitly setting phoenix.upsert.batch.size property in my oozie workflow then my job ran successfully.
By default, phoenix.upsert.batch.size is 1000. Hence, the commits were failing with a huge batch size of 1000. Thanks, Anil Gupta On Sun, Feb 14, 2016 at 8:03 PM, Heng Chen <heng.chen.1...@gmail.com> wrote: > I am not sure whether "upsert batch size in phoenix" equals HBase Client > batch puts size or not. > > But as log shows, it seems there are 2000 actions send to hbase one time. > > 2016-02-15 11:38 GMT+08:00 anil gupta <anilgupt...@gmail.com>: > >> My phoenix upsert batch size is 50. You mean to say that 50 is also a lot? >> >> However, AsyncProcess is complaining about 2000 actions. >> >> I tried with upsert batch size of 5 also. But it didnt help. >> >> On Sun, Feb 14, 2016 at 7:37 PM, anil gupta <anilgupt...@gmail.com> >> wrote: >> >> > My phoenix upsert batch size is 50. You mean to say that 50 is also a >> lot? >> > >> > However, AsyncProcess is complaining about 2000 actions. >> > >> > I tried with upsert batch size of 5 also. But it didnt help. >> > >> > >> > On Sun, Feb 14, 2016 at 6:43 PM, Heng Chen <heng.chen.1...@gmail.com> >> > wrote: >> > >> >> 2016-02-14 12:34:23,593 INFO [main] >> >> org.apache.hadoop.hbase.client.AsyncProcess: #1, waiting for 2000 >> >> actions to finish >> >> >> >> It means your writes are too many, please decrease the batch size of >> your >> >> puts, and balance your requests on each RS. >> >> >> >> 2016-02-15 4:53 GMT+08:00 anil gupta <anilgupt...@gmail.com>: >> >> >> >> > After a while we also get this error: >> >> > 2016-02-14 12:45:10,515 WARN [main] >> >> > org.apache.phoenix.execute.MutationState: Swallowing exception and >> >> > retrying after clearing meta cache on connection. >> >> > java.sql.SQLException: ERROR 2008 (INT10): Unable to find cached >> index >> >> > metadata. ERROR 2008 (INT10): ERROR 2008 (INT10): Unable to find >> >> > cached index metadata. key=-594230549321118802 >> >> > region=BI.SALES,,1455470578449.44e39179789041b5a8c03316730260e7. >> Index >> >> > update failed >> >> > >> >> > We have already set: >> >> > >> >> > >> >> >> <name>phoenix.coprocessor.maxServerCacheTimeToLiveMs</name><value>180000</value> >> >> > >> >> > Upset batch size is 50. Write are quite frequent so the cache would >> >> > not timeout in 180000ms >> >> > >> >> > >> >> > On Sun, Feb 14, 2016 at 12:44 PM, anil gupta <anilgupt...@gmail.com> >> >> > wrote: >> >> > >> >> > > Hi, >> >> > > >> >> > > We are using phoenix4.4, hbase 1.1(hdp2.3.4). >> >> > > I have a MR job that is using PhoenixOutputFormat. My job keeps on >> >> > failing >> >> > > due to following error: >> >> > > >> >> > > 2016-02-14 12:29:43,182 INFO [main] >> >> > org.apache.hadoop.hbase.client.AsyncProcess: #1, waiting for 2000 >> >> actions >> >> > to finish >> >> > > 2016-02-14 12:29:53,197 INFO [main] >> >> > org.apache.hadoop.hbase.client.AsyncProcess: #1, waiting for 2000 >> >> actions >> >> > to finish >> >> > > 2016-02-14 12:30:03,212 INFO [main] >> >> > org.apache.hadoop.hbase.client.AsyncProcess: #1, waiting for 2000 >> >> actions >> >> > to finish >> >> > > 2016-02-14 12:30:13,225 INFO [main] >> >> > org.apache.hadoop.hbase.client.AsyncProcess: #1, waiting for 2000 >> >> actions >> >> > to finish >> >> > > 2016-02-14 12:30:23,239 INFO [main] >> >> > org.apache.hadoop.hbase.client.AsyncProcess: #1, waiting for 2000 >> >> actions >> >> > to finish >> >> > > 2016-02-14 12:30:33,253 INFO [main] >> >> > org.apache.hadoop.hbase.client.AsyncProcess: #1, waiting for 2000 >> >> actions >> >> > to finish >> >> > > 2016-02-14 12:30:43,266 INFO [main] >> >> > org.apache.hadoop.hbase.client.AsyncProcess: #1, waiting for 2000 >> >> actions >> >> > to finish >> >> > > 2016-02-14 12:30:53,279 INFO [main] >> >> > org.apache.hadoop.hbase.client.AsyncProcess: #1, waiting for 2000 >> >> actions >> >> > to finish >> >> > > 2016-02-14 12:31:03,293 INFO [main] >> >> > org.apache.hadoop.hbase.client.AsyncProcess: #1, waiting for 2000 >> >> actions >> >> > to finish >> >> > > 2016-02-14 12:31:13,305 INFO [main] >> >> > org.apache.hadoop.hbase.client.AsyncProcess: #1, waiting for 2000 >> >> actions >> >> > to finish >> >> > > 2016-02-14 12:31:23,318 INFO [main] >> >> > org.apache.hadoop.hbase.client.AsyncProcess: #1, waiting for 2000 >> >> actions >> >> > to finish >> >> > > 2016-02-14 12:31:33,331 INFO [main] >> >> > org.apache.hadoop.hbase.client.AsyncProcess: #1, waiting for 2000 >> >> actions >> >> > to finish >> >> > > 2016-02-14 12:31:43,345 INFO [main] >> >> > org.apache.hadoop.hbase.client.AsyncProcess: #1, waiting for 2000 >> >> actions >> >> > to finish >> >> > > 2016-02-14 12:31:53,358 INFO [main] >> >> > org.apache.hadoop.hbase.client.AsyncProcess: #1, waiting for 2000 >> >> actions >> >> > to finish >> >> > > 2016-02-14 12:32:03,371 INFO [main] >> >> > org.apache.hadoop.hbase.client.AsyncProcess: #1, waiting for 2000 >> >> actions >> >> > to finish >> >> > > 2016-02-14 12:32:13,385 INFO [main] >> >> > org.apache.hadoop.hbase.client.AsyncProcess: #1, waiting for 2000 >> >> actions >> >> > to finish >> >> > > 2016-02-14 12:32:23,399 INFO [main] >> >> > org.apache.hadoop.hbase.client.AsyncProcess: #1, waiting for 2000 >> >> actions >> >> > to finish >> >> > > 2016-02-14 12:32:33,412 INFO [main] >> >> > org.apache.hadoop.hbase.client.AsyncProcess: #1, waiting for 2000 >> >> actions >> >> > to finish >> >> > > 2016-02-14 12:32:43,428 INFO [main] >> >> > org.apache.hadoop.hbase.client.AsyncProcess: #1, waiting for 2000 >> >> actions >> >> > to finish >> >> > > 2016-02-14 12:32:53,443 INFO [main] >> >> > org.apache.hadoop.hbase.client.AsyncProcess: #1, waiting for 2000 >> >> actions >> >> > to finish >> >> > > 2016-02-14 12:33:03,457 INFO [main] >> >> > org.apache.hadoop.hbase.client.AsyncProcess: #1, waiting for 2000 >> >> actions >> >> > to finish >> >> > > 2016-02-14 12:33:13,472 INFO [main] >> >> > org.apache.hadoop.hbase.client.AsyncProcess: #1, waiting for 2000 >> >> actions >> >> > to finish >> >> > > 2016-02-14 12:33:23,486 INFO [main] >> >> > org.apache.hadoop.hbase.client.AsyncProcess: #1, waiting for 2000 >> >> actions >> >> > to finish >> >> > > 2016-02-14 12:33:33,524 INFO [main] >> >> > org.apache.hadoop.hbase.client.AsyncProcess: #1, waiting for 2000 >> >> actions >> >> > to finish >> >> > > 2016-02-14 12:33:43,538 INFO [main] >> >> > org.apache.hadoop.hbase.client.AsyncProcess: #1, waiting for 2000 >> >> actions >> >> > to finish >> >> > > 2016-02-14 12:33:53,551 INFO [main] >> >> > org.apache.hadoop.hbase.client.AsyncProcess: #1, waiting for 2000 >> >> actions >> >> > to finish >> >> > > 2016-02-14 12:34:03,565 INFO [main] >> >> > org.apache.hadoop.hbase.client.AsyncProcess: #1, waiting for 2000 >> >> actions >> >> > to finish >> >> > > 2016-02-14 12:34:03,953 INFO >> [hconnection-0xe82ca6e-shared--pool2-t16] >> >> > org.apache.hadoop.hbase.client.AsyncProcess: #1, table=BI.SALES, >> >> > attempt=10/35 failed=2000ops, last exception: null on >> hdp3.truecar.com >> >> ,16020,1455326291512, >> >> > tracking started null, retrying after=10086ms, replay=2000ops >> >> > > 2016-02-14 12:34:13,578 INFO [main] >> >> > org.apache.hadoop.hbase.client.AsyncProcess: #1, waiting for 2000 >> >> actions >> >> > to finish >> >> > > 2016-02-14 12:34:23,593 INFO [main] >> >> > org.apache.hadoop.hbase.client.AsyncProcess: #1, waiting for 2000 >> >> actions >> >> > to finish >> >> > > >> >> > > I have never seen anything like this. Can anyone give me pointers >> >> about >> >> > > this problem? >> >> > > >> >> > > -- >> >> > > Thanks & Regards, >> >> > > Anil Gupta >> >> > > >> >> > >> >> > >> >> > >> >> > -- >> >> > Thanks & Regards, >> >> > Anil Gupta >> >> > >> >> >> > >> > >> > >> > -- >> > Thanks & Regards, >> > Anil Gupta >> > >> >> >> >> -- >> Thanks & Regards, >> Anil Gupta >> > > -- Thanks & Regards, Anil Gupta
