Deleting entire buckets
Hi, What is the recommended way of deleting all of the keys in a bucket? Is there a way to delete all the keys without impacting the read performance of the cluster, like a low priority clean up job? I found the following script so far (I guess running it at off peak would be solving the priority problem): http://stackoverflow.com/a/14531820/127508 Is this good, bad approach? Thanks in advance, Istvan -- the sun shines for all ___ riak-users mailing list riak-users@lists.basho.com http://lists.basho.com/mailman/listinfo/riak-users_lists.basho.com
Re: Deleting entire buckets
Istvan, This can be performed using the method described in that comment by Randall on stackoverflow, which performs a keylisting, or via a MapReduce job. I recommend doing this type of administrative work during off peak hours regardless of whether you go the keylisting or MapReduce route. It's also important to keep in mind that if you're performing these deletes to reclaim disk space, you will likely increase disk usage in the short term. Both Bitcask and LevelDB are append only backends, with merging / compaction processes that are triggered in the background. Thanks, John Caprice On Tue, Feb 25, 2014 at 1:57 PM, István wrote: > Hi, > > What is the recommended way of deleting all of the keys in a bucket? > > Is there a way to delete all the keys without impacting the read > performance of the cluster, like a low priority clean up job? > > I found the following script so far (I guess running it at off peak > would be solving the priority problem): > > http://stackoverflow.com/a/14531820/127508 > > Is this good, bad approach? > > Thanks in advance, > Istvan > > -- > the sun shines for all > > ___ > riak-users mailing list > riak-users@lists.basho.com > http://lists.basho.com/mailman/listinfo/riak-users_lists.basho.com > ___ riak-users mailing list riak-users@lists.basho.com http://lists.basho.com/mailman/listinfo/riak-users_lists.basho.com
Re: Deleting entire buckets
Hey John, Thank you for the answer. Yes, my plan was to run outside peak hours. I was not sure if listing the keys is better than MapReduce in terms of controllable additional load on the cluster, so I guess I need to simple measure it. Thanks, I. On Tue, Feb 25, 2014 at 12:18 PM, John Caprice wrote: > Istvan, > > This can be performed using the method described in that comment by Randall > on stackoverflow, which performs a keylisting, or via a MapReduce job. I > recommend doing this type of administrative work during off peak hours > regardless of whether you go the keylisting or MapReduce route. > > It's also important to keep in mind that if you're performing these deletes > to reclaim disk space, you will likely increase disk usage in the short > term. Both Bitcask and LevelDB are append only backends, with merging / > compaction processes that are triggered in the background. > > Thanks, > > John Caprice > > > On Tue, Feb 25, 2014 at 1:57 PM, István wrote: >> >> Hi, >> >> What is the recommended way of deleting all of the keys in a bucket? >> >> Is there a way to delete all the keys without impacting the read >> performance of the cluster, like a low priority clean up job? >> >> I found the following script so far (I guess running it at off peak >> would be solving the priority problem): >> >> http://stackoverflow.com/a/14531820/127508 >> >> Is this good, bad approach? >> >> Thanks in advance, >> Istvan >> >> -- >> the sun shines for all >> >> ___ >> riak-users mailing list >> riak-users@lists.basho.com >> http://lists.basho.com/mailman/listinfo/riak-users_lists.basho.com > > -- the sun shines for all ___ riak-users mailing list riak-users@lists.basho.com http://lists.basho.com/mailman/listinfo/riak-users_lists.basho.com
Re: [ANN] Riak 1.4.8
Hi, After upgrading to 1.4.8 in staging, last week, we've been seeing a fairly constant extra amount of riak hits in the stats. Consistently 300-400 a minute. Can get lost in the much higher, varying levels during busy hours, but outside of that it's quite visible as a baseline. The logs show a lot of AAE work going on. I wondered if this was normal behaviour of Riak cleaning up the AAE-related bug from earlier versions, and will go away eventually? Or that unlikely? I emphasize that there's no significant server load resulting from this -- I'm just curious if it's something I need to look into or not. Cheers, Toby On 21 February 2014 04:00, Tom Santero wrote: > Hi, > > Today, Basho released a minor update of Riak and Riak Enterprise Edition, > version 1.4.8. This bug-fix release addresses a regression[0] in Riak's > active anti-entropy (AAE) system. The regression existed in versions 1.4.3 > through 1.4.7. > > The complete release notes[1] and package downloads[2] are now available. > > Thanks for being the best community ever. > > Regards, > The Basho Team > > [0] > http://lists.basho.com/pipermail/riak-users_lists.basho.com/2014-January/014551.html > [1] https://github.com/basho/riak/blob/1.4/RELEASE-NOTES.md > [2] http://docs.basho.com/riak/latest/downloads/ > > > ___ > riak-users mailing list > riak-users@lists.basho.com > http://lists.basho.com/mailman/listinfo/riak-users_lists.basho.com > -- Turning and turning in the widening gyre The falcon cannot hear the falconer Things fall apart; the center cannot hold Mere anarchy is loosed upon the world ___ riak-users mailing list riak-users@lists.basho.com http://lists.basho.com/mailman/listinfo/riak-users_lists.basho.com
