Re: Hash join confusion

Not sure if it's related, coz your DDL does not have DESC columns, but we do have a sort-merge-join bug fix in 4.8.0: https://issues.apache.org/jira/browse/PHOENIX-2894. Otherwise could you please just file a JIRA and assign to me? Thanks a lot! Thanks, Maryann On Tue, Oct 4, 2016 at 8:24 PM, S

Re: Hash join confusion

Hi Maryann, Here are the 2 DDLs (for data and index tables): CREATE TABLE IF NOT EXISTS "ldmns:exDocStoreb" (CURRENT_TIMESTAMP BIGINT NOT NULL, ID VARCHAR(96), BINARY_CURR_EXDOC VARBINARY, CURR_CHECKSUM VARCHAR(32), BINARY_PREV_EXDOC VARBINARY, PREV_CHECKSUM VARCHAR(32), PREV_TIMESTAMP BIGINT, S

Re: Jdbc secure connection -- exception

Hi Vikram, See https://issues.apache.org/jira/browse/PHOENIX-1754 This is presently an open issue. You will not be able to use the convenience "Kerberos login via URL" when on Windows. You will need to manually perform your Kerberos login (via JAAS or Hadoop's UserGroupInformation class) and

Re: Question regarding designing row keys

Sure, but he should create the Phoenix table already salted. În Mar, 4 oct. 2016, 17:21 Mujtaba Chohan, a scris: > If you lead with timestamp key, you might want to consider experimenting > with salting as writing data > would hotspot on single region if k

Jdbc secure connection -- exception

Hi, I am trying to connect to phoenix using jdbc with Kerberos from windows DriverManager.getConnection("jdbc:phoenix::2181:/hbase:C:\\kbfile\\xyz.keytab:x...@yyy.com", props); And getting the following exception Malformed connection url. : jdbc:phoenix::2181:/hbase:C:\kbfile\xyz.keyta

Jdbc secure connection -- exception

Hi, I am trying to connect to phoenix using jdbc with Kerberos from windows DriverManager.getConnection("jdbc:phoenix::2181:/hbase:C:\\kbfile\\xyz.keytab:x...@yyy.com", props); And getting the following exception Malformed connection url. : jdbc:phoenix::2181:/hbase:C:\kbfile\xyz.keyt

Re: Hash join confusion

Thanks Maryann. I will share the details in a few hours. Under heavy load scenario, the default hash join failed with time-out (and memory issue), so I switched to sort-merge. But sort-merge is missing data randomly. So, as of now I am not sure what is the issue with sort-merge join. Hash join do

Re: Hash join confusion

Hi Sumit, Thank you for the update! Would you mind sharing the queries and their plans, as well as the DDL for both the data tables and the index? And just to confirm, you are saying hash joins are working, is it with changes to the config or without? Thanks, Maryann On Tue, Oct 4, 2016 at 9:17

Re: Hash join confusion

Thank you Maryann. >From the time I have moved to sort-merge join, my use cases have stopped >working. However, if I remove the hint (and fall back to hash), then they all >work. I am on phoenix 4.6/ hbase 1.1 I thought just changing the join algorithm would be enough. I would assume that changi

Re: Question regarding designing row keys

If you lead with timestamp key, you might want to consider experimenting with salting as writing data would hotspot on single region if keys are monotonically increasing. On Tue, Oct 4, 2016 at 8:04 AM, Ciureanu Constantin < ciureanu.constan...@gmail.com> wr

Re: Question regarding designing row keys

select * from metric_table where metric_type='x' -- so far so good and timestamp > 'start_date' and timestamp < 'end_date'. -- here in case the timestamp is long (BIGINT in Phoenix) - it should work fine! Try also with "timestamp between (x and y)" Anyway - my proposal would be to reverse the key