Member since
07-08-2013
60
Posts
8
Kudos Received
15
Solutions
My Accepted Solutions
Title | Views | Posted |
---|---|---|
4596 | 09-11-2018 09:29 AM | |
1539 | 02-15-2018 09:03 AM | |
5107 | 08-28-2017 09:15 AM | |
4213 | 07-21-2017 08:40 AM | |
12391 | 06-23-2017 08:51 AM |
09-11-2018
09:29 AM
1 Kudo
Hi, This was missed in the docs, I opened an internal ticket to have it added to that page. We did the 6.0 certification on RHEL-compatible 7.5 OSes back in June. Sorry about this! J-D
... View more
04-19-2018
04:17 PM
It's hard to tell from just this log snippet if anything really bad is happening, but for sure you have a tablet server that seems to be gone (192.168.1.XX:7050: connect: Connection refused) during that time. Is ksck returning all ok? Do you have a range partition that covers the time range you are writing into?
... View more
03-16-2018
12:34 PM
Hi, This looks like a different flavor of something we encountered while testing Sentry in 5.13 (that we didn't resolve). I pinged a few folks internally. FWIW I tried just spinning up a secure CM5.14.1 + CDH5.12.2 cluster with all the services and my HMS came up just fine without that config. Can you describe more how you upgraded CM, if any other configs were changed before/after the upgrade, and if you added any services before or after? As an aside, the Impala forum is probably not the best place to get help for this issue. Thanks, J-D
... View more
02-15-2018
09:03 AM
1 Kudo
Hi Rajesh, Yes that's what it means. J-D
... View more
08-28-2017
09:15 AM
"PARTITIONS 1" means that you are trying to create a table with 1 hash bucket, you need at least 2 (like the error you're getting says). This is a way to force you to think about how you want to distribute your data. Only 2 tservers won't work with the default replication factor of 3, as per http://kudu.apache.org/docs/known_issues.html#_tables
... View more
08-28-2017
09:03 AM
@hindog wrote: "Nor the browser column" -- What if the "browser" column WAS the primary key? Would increment be possible in similar fashion to upsert, the difference being that internally Kudu would read the existing value and add it to the increment value before saving? The other Kudu limitation that this runs into is that row key columns cannot be modified. It would be somewhat easier to implement atomic increments on non-key columns, but then that means doing a lot of updates and it's not something Kudu excels at e.g. it supports modifying data but the higher the rate the slower the reads are.
... View more
07-31-2017
09:21 AM
Kudu doesn't do anything special to resolve hostnames, so it should use the hosts file by way of doing the resolving. You can also see a 5s delay when trying to resolve the address, I would recommend googling "dns 5 seconds delay" and debug your network situation from there.
... View more
07-21-2017
09:36 AM
Deletes like this should be fine, even more ideal when you need to delete huge batches of rows is to hopefully have range partitions setup so that you can just drop them (see this doc), but for millions of rows it shouldn't timeout unless there's something misconfigured.
... View more
07-21-2017
09:16 AM
Dropping the table and recreating it would be a lot more efficient than deleting every row one by one after reading them (which is what "delete from" does). All the logs that pertain to tablet 871c8123905c4e529a233c18751f8154 for that tablet server could be useful. The full INFO file would be perfect if you can zip it and host it on a server somewhere.
... View more
07-21-2017
08:40 AM
1 Kudo
You need to run ksck as the Kudu superuser in order to get the full functionality, this is defined by "--superuser_acl" and it defaults to the user that runs Kudu which is normally "kudu".
... View more