Re: Hot, large row

2014-07-28 Thread Keith Wright
@cassandra.apache.org>" mailto:user@cassandra.apache.org>> Subject: Re: Hot, large row Is it the accumulated tombstones on a row that make it act as if “wide”? Does cfhistograms count the tombstones or subtract them when reporting on cell-count for rows? (I don’t know.) -- Jack Krupansky Fr

Re: Hot, large row

2014-07-25 Thread Jack Krupansky
: Don Jackson Subject: Re: Hot, large row Ha, check out who filed that ticket! Yes I’m aware of it. My hope is that it was mostly addressed in CASSANDRA-6563 so I may upgrade from 2.0.6 to 2.0.9. I’m really just surprised that others are not doing similar actions as I and thus experiencing

Re: Hot, large row

2014-07-25 Thread Keith Wright
gt;" mailto:user@cassandra.apache.org>> Date: Friday, July 25, 2014 at 10:06 AM To: "user@cassandra.apache.org<mailto:user@cassandra.apache.org>" mailto:user@cassandra.apache.org>> Cc: Don Jackson mailto:djack...@nanigans.com>> Subject: Re: Hot, large row https://issues.apache.org/jira/browse/CASSANDRA-6654

Re: Hot, large row

2014-07-25 Thread Ken Hancock
ue? > > I am hoping to upgrade to 2.0.9 which has improvements to remove > tombstones. > > From: Keith Wright > Reply-To: "user@cassandra.apache.org" > Date: Thursday, July 24, 2014 at 4:50 PM > To: "user@cassandra.apache.org" > Cc: Don Jackson > &g

Re: Hot, large row

2014-07-25 Thread DuyHai Doan
Hello Keith 1. Periodically seeing one node stuck in CMS GC causing high read latency. Seems to recover on its own after an hour or so How many nodes do you have ? And how many distinct user_id roughtly is there ? Looking at your jvm settings it seems that you have the GC log enabled. It

Re: Hot, large row

2014-07-25 Thread Duncan Sands
Hi Keith, On 25/07/14 14:43, Keith Wright wrote: Answers to your questions below but in the end I believe the root issue here is that LCS is clearly not compacting away as it should resulting in reads across many SSTables which as you noted is “fishy”. I’m considering filing a JIRA for this, s

Re: Hot, large row

2014-07-25 Thread Keith Wright
@cassandra.apache.org<mailto:user@cassandra.apache.org>" mailto:user@cassandra.apache.org>> Subject: Re: Hot, large row What are your jvm settings? Your read pattern implies that you may fetch lots of data into memory (reading all skus for a given user), maybe it stressed too much the jvm.

Re: Hot, large row

2014-07-24 Thread DuyHai Doan
emove > tombstones. > > From: Keith Wright > Reply-To: "user@cassandra.apache.org" > Date: Thursday, July 24, 2014 at 4:50 PM > To: "user@cassandra.apache.org" > Cc: Don Jackson > Subject: Re: Hot, large row > > When a node is showing the high CMS

Re: Hot, large row

2014-07-24 Thread Keith Wright
gt;> Date: Thursday, July 24, 2014 at 4:50 PM To: "user@cassandra.apache.org<mailto:user@cassandra.apache.org>" mailto:user@cassandra.apache.org>> Cc: Don Jackson mailto:djack...@nanigans.com>> Subject: Re: Hot, large row When a node is showing the high CMS issue, io is

Re: Hot, large row

2014-07-24 Thread Keith Wright
rsday, July 24, 2014 at 4:44 PM To: "user@cassandra.apache.org<mailto:user@cassandra.apache.org>" mailto:user@cassandra.apache.org>> Subject: Re: Hot, large row For global_user_event_skus_v2 1. number of SSTables per read is quite huge. Considering you're using LCS, i

Re: Hot, large row

2014-07-24 Thread DuyHai Doan
1597 us: 177538 > >1916 us: 169568 > >2299 us: 177617 > >2759 us: 202552 > >3311 us: 198101 > >3973 us: 179807 > >4768 us: 159149 > >5722 us: 142282 > >6866 us: 126124 > >8239 us: 111529 > >9887

Re: Hot, large row

2014-07-24 Thread Keith Wright
cells: 48 642 cells: 25 770 cells: 9 924 cells: 9 1109 cells: 4 1331 cells: 3 1597 cells: 1 1916 cells: 2 From: Jack Krupansky mailto:j...@basetechnology.com>> Reply-To: "user@cassandra.apache.org<mailto:user@cassandra.apache.org>" mailto:user@cassandra.apache.org&g

Re: Hot, large row

2014-07-24 Thread Jack Krupansky
? -- Jack Krupansky From: DuyHai Doan Sent: Thursday, July 24, 2014 3:53 PM To: user@cassandra.apache.org Subject: Re: Hot, large row Your extract of cfhistograms show that there are no particular "wide rows". The widest has 61214 cells which is big but not that huge to be really

Re: Hot, large row

2014-07-24 Thread DuyHai Doan
e a > large row? If it were a small row being called repeatedly, I assume OS/key > cache would make that a VERY fast operation. > > Thanks > > From: DuyHai Doan > Reply-To: "user@cassandra.apache.org" > Date: Thursday, July 24, 2014 at 3:53 PM > > To: "u

Re: Hot, large row

2014-07-24 Thread Keith Wright
lt;mailto:user@cassandra.apache.org>" mailto:user@cassandra.apache.org>> Date: Thursday, July 24, 2014 at 3:53 PM To: "user@cassandra.apache.org<mailto:user@cassandra.apache.org>" mailto:user@cassandra.apache.org>> Subject: Re: Hot, large row Your extract of cf

Re: Hot, large row

2014-07-24 Thread DuyHai Doan
8 > > 1916 cells: 773 > > 2299 cells: 495 > > 2759 cells: 268 > > 3311 cells: 150 > > 3973 cells: 100 > > 4768 cells: 42 > > 5722 cells: 24 > > 6866 cells: 12 > > 8239 cells: 9 > > 9887 cells: 3 > > 11864 ce

Re: Hot, large row

2014-07-24 Thread Keith Wright
e.org>> Date: Thursday, July 24, 2014 at 3:01 PM To: "user@cassandra.apache.org<mailto:user@cassandra.apache.org>" mailto:user@cassandra.apache.org>> Subject: Re: Hot, large row "How can I detect wide rows?" --> nodetool cfhistograms Look at column "

Re: Hot, large row

2014-07-24 Thread DuyHai Doan
"How can I detect wide rows?" --> nodetool cfhistograms Look at column "Column count" (last column) and identify a line in this column with very high value of "Offset". In a well designed application you should have a gaussian distribution where 80% of your row have a similar number of columns.

Hot, large row

2014-07-24 Thread Keith Wright
Hi all, We are seeing an issue where basically daily one of our nodes spikes in load and is churning in CMS heap pressure. It appears that reads are backing up and my guess is that our application is reading a large row repeatedly. Our write structure can lead itself to wide rows very infr