i ran a simple mysqlslap today about 2 and a half hours ago now my server wont respond to a simple q command also it doesn't seem to be responding to a control c command as well. however my website that is on same sever responds to requests. however my web-admin tool does not but i assume this is because it uses Linux to authenticate and Linux probably does not have any cpu time to give to it to authenticate.also whene i try to log in to my website i get a :
Bad Request

Your browser sent a request that this server could not understand.

the logon process uses mysql.

i thinking i probably have caused a denial of service on the system.

my top command output seems to be frozen with this at the top of the list

22296 user 20 0 994m 51m 92 S 2 20.6 4:03.48 mysqlslap
836 user 20 0 219m 9448 288 S 1 3.7 13:44.66 mysqld
33 user 20 0 0 0 0 D 1 0.0 0:36.76 kswapd0
23122 user 20 0 17304 5052 1152 D 0 2.0 0:03.22 monitor.pl
949 user 20 0 47188 616 356 S 0 0.2 0:22.33 postgres
23369 user 20 0 14156 3952 952 D 0 1.5 0:00.16 perl
23361 user 20 0 14156 4036 940 D 0 1.6 0:00.15 perl
23223 user 20 0 12316 3568 1028 D 0 1.4 0:01.92 monitor.pl
23257 user 20 0 11660 4016 968 D 0 1.6 0:01.63 monitor.pl
23265 user 20 0 15624 2464 716 D 0 1.0 0:01.36 perl
23365 user 20 0 11264 6260 1368 D 0 2.4 0:00.29 monitor.pl
21 user 20 0 0 0 0 S 0 0.0 0:24.66 kblockd/0
22 user 20 0 0 0 0 S 0 0.0 0:25.01 kblockd/1
22776 user 20 0 2620 544 316 R 0 0.2 0:07.29 top
23351 user 20 0 15212 6212 928 D 0 2.4 0:00.43 perl
23367 user 20 0 13884 2116 736 D 0 0.8 0:00.04 perl
23386 user 20 0 14016 2196 868 D 0 0.9 0:00.03 perl
947 user 20 0 47052 40 0 S 0 0.0 0:40.53 postgres
23310 user 20 0 11264 3348 1104 S 0 1.3 0:00.79 monitor.pl
23339 user 20 0 11264 5764 1320 S 0 2.2 0:00.52 monitor.pl
23370 user 20 0 13884 2240 736 D 0 0.9 0:00.04 perl
23371 user 20 0 13884 2224 736 D 0 0.9 0:00.04 perl
23373 user 20 0 13884 1968 696 D 0 0.8 0:00.04 perl
23374 user 20 0 13884 2380 736 D 0 0.9 0:00.04 perl
23388 user 20 0 2628 764 572 D 0 0.3 0:00.02 cron

if anyone could help with maby explain this or a way to fix it with out having to reboot my server i have both the windows open for top and the script that ran the mysqlslap

Are you running this load-test on a production server?

1 members found this post helpful.

yes this is on the production server.should note that the load test is at the last 2 test which are testing 4 times our expected load. OK looks like the test just failed it came back with

./stsql.sh: line 22: 22296 Segmentation fault

mysqlslap --user=*******--auto-generate-sql --concurrency=100 --number-of-queries=30000 --password=********** >> 30000.txt

how ever it starting the last test

CPU load averages 142.85 (1 min) 139.60 (5 mins) 131.53 (15 mins)
CPU usage 7% user, 6% kernel, 87% IO, 0% idle
Virtual memory 884 MB total, 329.22 MB used

Unsure of how to cancel the test,.. but it should eventually finish.

1 members found this post helpful.

test just failed and i had a safe guard in the script that after each test it would sleep for 5 minutes so gave me a chance to end the process. can you maby explain why so many processes went in to a D state. also is there a way to clear the virtual memory seems theirs still alot in it

Without testing this myself, I cant say why your processes started sleeping.

I will say this though: In the future, -do not- run a load test on a production server

If you need to do something like that, you need to run it on a development version far, far away from your production environment. Even if that means running a Virtual Machine.

1 members found this post helpful.

we ran this on development server already and then was asked to test the load on production to make sure it could still handle the load after the changes where implemented even against all advice we had given theme. normally i would never ever run a load like that on production but higher ups had asked for it so.