Re: performance issue involving "insert as select"

2016-12-11 Thread Todd Lipcon
Hi Rotem, On Thu, Dec 8, 2016 at 3:25 AM, Rotem Gabay wrote: > Hi, I have a small cluster on which I tried to run some performance tests > on kudu, In order to populate some data I have made simple "insert as > select" from simple HDFS table that took 10 minutes to finish. I then tried > to dup

Re: Hints about encoding and compression

2016-12-11 Thread Todd Lipcon
Hi Nicolas, Apologies for the slow response. Answers inline: On Mon, Dec 5, 2016 at 8:21 PM, Nicolas Fouché wrote: > Hi, > > I'm evaluating Kudu and I'd need some hints about column encoding and > compression. > > A- Does it make sense adding LZ4 compression to a field with Dictionary > Encodin

Re: Kudu on kerberos enabled cluster

2016-12-11 Thread Todd Lipcon
On Sun, Dec 4, 2016 at 11:40 PM, Amit Adhau wrote: > Thank you Todd, > > In case of, using non-Kerberized Kudu alongside Kerberized > Impala/HDFS/etc - Are you saying that we can have kudu masters and kudu > tablets outside of the kerberos enabled HDFS instances, but then I guess > kudu will not

Re: Good way to find "Real" size of the tables

2016-12-11 Thread Todd Lipcon
Hey Rick, Just wanted to check and see if you were able to make any progress on the experiments you were running. Would be great to share your findings or any issues you encountered. -Todd On Thu, Dec 1, 2016 at 10:49 PM, Weber, Richard wrote: > Comments below > > > On Nov 30, 2016, at 4:29 PM

Re: About data file size and on-disk size

2016-12-11 Thread Todd Lipcon
Just a follow-up note here: if you did end up cherry-picking that change, you should also be sure to cherry-pick faa587c639aa9e5dcf3fac04259f46ba1921140a to avoid a potential data loss bug. On Wed, Nov 30, 2016 at 9:00 AM, Adar Dembo wrote: > If you're comfortable rebuilding Kudu from source, yo