Hello Everyone,
I installed lzo in our cluster. And I found that a insert job which
normally takes 2 minutes will take 20 minutes now.
The following is part of vmstat logs on the working regionserver when
the insert job is running:
procs -----------memory---------- ---swap-- -----io---- --system--
-----cpu------
r b swpd free buff cache si so bi bo in cs us sy id
wa st
1 0 2300556 651964 13792 1399784 0 0 0 0 1533 22829 10 7
83 0 0
1 0 2300460 651360 13792 1400312 100 0 100 0 1144 25314 11 8
81 0 0
1 0 2300460 650616 13792 1400888 0 0 0 168 1574 24416 10 8
82 0 0
1 0 2300444 650036 13792 1401252 36 0 36 0 1541 24622 10 8
82 0 0
1 0 2300440 649044 13800 1401788 0 0 0 3032 1166 22469 11 8
80 1 0
1 0 2300388 648764 13800 1402296 32 0 32 0 1540 20633 9 7
83 0 0
2 0 2300292 648316 13804 1402684 100 0 100 0 1124 24694 10 8
82 0 0
1 0 2300292 647612 13804 1403664 0 0 0 0 1568 25410 10 8
82 0 0
1 0 2300288 647076 13804 1404132 0 0 0 0 1132 25667 10 8
81 0 0
2 0 2300284 646060 13812 1404804 32 0 32 3044 1563 23355 10 8
81 1 0
1 0 2300232 645588 13812 1405500 32 0 32 0 1140 24416 11 8
81 0 0
2 0 2300124 645116 13812 1405856 132 0 132 0 1588 24845 10 8
82 0 0
The following is result of "top":
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+
COMMAND
28603 root 24 0 1549m 433m 11m S 91.1 11.0 64:27.35
java
17979 root 24 0 1555m 179m 9384 S 54.2 4.6 11:33.44
java
5284 root 23 0 918m 260m 11m S 3.7 6.6 0:48.24
java
19007 root 18 0 4502m 764m 10m S 2.3 19.4 20:02.20 java
and the first two are "dadtanode" and "regionserver".
I'm sorry there is no more infomation about this I can give. Anyone has any
experience about this?
Thanks a lot.
LvZheng.
