Re: Kudu Data Storage Size Mistmatch On dashboard / data folder

2016-04-28 Thread Amit Adhau
Thank you Todd. Thanks, Amit On Thu, Apr 28, 2016 at 9:25 PM, Todd Lipcon wrote: > Hi Amit, > > What you're probably seeing is container pre-allocation. We use > 'fallocate' to preallocate space in the block container files, to avoid > fragmentation on disk. So, the disk usage will increase in

Re: Weekly update 4/25

2016-04-28 Thread Mike Percy
Thanks for filing it, Jordan. Great writeup too. Mike On Thu, Apr 28, 2016 at 12:54 PM, Jordan Birdsell < jordan.birdsell.k...@statefarm.com> wrote: > Opened KUDU-1431 > > > > *From:* Mike Percy [mailto:mpe...@apache.org] > *Sent:* Thursday, Apri

RE: Weekly update 4/25

2016-04-28 Thread Jordan Birdsell
Opened KUDU-1431 From: Mike Percy [mailto:mpe...@apache.org] Sent: Thursday, April 28, 2016 1:55 PM To: user@kudu.incubator.apache.org Subject: Re: Weekly update 4/25 Hey Jordan, It would definitely be helpful if you could file a JIRA to track this

Re: Weekly update 4/25

2016-04-28 Thread Mike Percy
Hey Jordan, It would definitely be helpful if you could file a JIRA to track this. The initial version of tablet history GC that I am currently working on as part of KUDU-236 won't yet support this type of SLA-based removal, since the current changes are much simpler than that since they are more

RE: Weekly update 4/25

2016-04-28 Thread Jordan Birdsell
Todd, Should a JIRA be opened to track this? From: Jordan Birdsell Sent: Tuesday, April 26, 2016 2:07 PM To: user@kudu.incubator.apache.org Subject: RE: Weekly update 4/25 Today we solve this on an RDBMS (DB2) platform, however when data is replicated to the cluster, we need to be able to addre

Re: Kudu Data Storage Size Mistmatch On dashboard / data folder

2016-04-28 Thread Todd Lipcon
Hi Amit, What you're probably seeing is container pre-allocation. We use 'fallocate' to preallocate space in the block container files, to avoid fragmentation on disk. So, the disk usage will increase in multiples of 32MB within each data container file. You can see the number of active containers

Kudu Data Storage Size Mistmatch On dashboard / data folder

2016-04-28 Thread Amit Adhau
Hi Kudu team, I have below observations for kudu data storage size mismatch; I was having similar observation last week when total on-disk size was between 4-5GB and data folder was showing 180GB. Hence, we have cleaned all kudu data and created new master and tablet data directories. That means