how many keys are you fetching? how many columns for each key?
On Thu, Jan 14, 2010 at 1:49 AM, Suhail Doshi suh...@mixpanel.com wrote:
I've been seeing multiget_slice take an extremely long time:
2010-01-14 07:44:00,513 INFO -- Cassandra, delay:
3.64020800591
Right now it's ~5-10 keys, with 5 columns per key.
Later it will be 64 keys (max) with 200 columns per key worst case.
Suhail
On Thu, Jan 14, 2010 at 9:45 AM, Jonathan Ellis jbel...@gmail.com wrote:
how many keys are you fetching? how many columns for each key?
On Thu, Jan 14, 2010 at 1:49
But as you can see value per column is just a byte so the time at which
it de-serializes the column shouldn't be horrid. Hopefully that's the right
thinking
On Thu, Jan 14, 2010 at 10:06 AM, Suhail Doshi digitalwarf...@gmail.comwrote:
Right now it's ~5-10 keys, with 5 columns per key.
Later
Looking at my data directory: 14 G. Just Index.db based files: 4.5 G.
Yes only one node so far.
vmstat -n 1 -S m
procs ---memory-- ---swap-- -io -system--
cpu
r b swpd free buff cache si sobibo in cs us sy id
wa
0 0 22585 32
it sounds like you just don't have enough ram for the OS to cache your
hot data set so you are getting killed on disk seeks. iostat -x 5
(for example) during load should verify this.
On Thu, Jan 14, 2010 at 11:19 AM, Suhail Doshi digitalwarf...@gmail.com wrote:
Looking at my data directory: 14
Yeah, I think you're right:
Device: rrqm/s wrqm/s r/s w/s rsec/s wsec/s avgrq-sz
avgqu-sz await svctm %util
sda1 34.0070.00 409.60 11.20 22596.80 649.6055.24
32.61 77.67 2.38 100.00
sda2 0.00 0.000.000.00 0.00