Details
-
Bug
-
Resolution: Incomplete
-
1.6.0 beta1
-
None
-
Operating System: All
Platform: All
Description
This seems quite basic load test on disk > ram, and quite easy to reproduce. Randi have you run such test so far??
1. I installed the latest version: 1.6.0beta2rc1_50_g6d70697
2. Run ./memcachetest -h 10.211.55.5 -i 10000000 (long load test)
3. Watched the minute stats which increased the load.
4. Ram got to the maximum on quota on this vm (~840M)
5. Disk writes peaked at 361K (according to the graph).
6. On the test program I start seeing a lot of the following error :
missing data for <06226345>
missing data for <03077567>
missing data for <02659817>
missing data for <03144154>
missing data for <06829352>
missing data for <09490985>
missing data for <03021173>
...
7. The VM became extremely slow
8. Here is the top command:
top - 23:37:15 up 7:11, 3 users, load average: 7.13, 5.63, 3.36
Tasks: 127 total, 2 running, 123 sleeping, 0 stopped, 2 zombie
Cpu(s): 1.3%us, 0.6%sy, 0.0%ni, 0.0%id, 96.8%wa, 1.3%hi, 0.0%si, 0.0%st
Mem: 1026392k total, 1019444k used, 6948k free, 1408k buffers
Swap: 2064344k total, 923380k used, 1140964k free, 95596k cached
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
7548 northsca 15 0 1133m 765m 1004 S 1.7 76.4 1:59.36 memcached
2557 root 15 0 217m 6288 1836 S 0.3 0.6 0:15.89 Xorg
9. Attached is the log file.
I don't expect misses on disk > RAM so starting this as Major P1, let me know if my understanding is incorrect (making rebalancing work is still higher priority though).