[c-nsp] mysql update

madunix madunix at gmail.com
Fri Jan 22 01:57:17 EST 2010


I have the following update procedure that update mySQL DB over the
internet between source Linux Centos (local machine on my net behind a
DMZ with real IP A.B.C.D) and target Linux fedora (web server
www.myweb.com) every day on a specific time 18:00 through a crontab on
my source linux server

server(source) ---DMZ---ASA---Router-----Internet----HostingCompany---Myweb(target)
[root at source]# mysql -u updatex -p -h www.myweb.com test < sample.SQL


[root at source]$ mysql -u updatex -p -h www.myweb.com test < sample.SQL
Enter password: *****
CURTIME()
19:41:44
CURTIME()
19:50:09

[root at source]$ mysql -u updatex -p -h www.myweb.com test < sample.SQL
Enter password:*****
CURTIME()
08:26:08
CURTIME()
08:26:34

I did the above procedure multiple times in different times in the
day. the duration of this procedure takes from 22sec to 10min
see above...., before a while it was running constant with duration of
30sec. I checked with my ISP, hosting company and network nothing been
changed from the structure/configuration.....

[root at source]# lsof -i -P | grep 3306
mysqld     3806   mysql   11u  IPv4  10926       TCP *:3306 (LISTEN)
mysql     15150    user    3u  IPv4 297528       TCP
192.168.10.5:8376->www.myweb.com:3306 (ESTABLISHED)

[root at target]# netstat -a |grep mysql
tcp        0      0 *:mysql                     *:*
     LISTEN
tcp        0      0 www.myweb.:mysql A.B.C.D:8366 TIME_WAIT
tcp        0     11 www.myweb.:mysql A.B.C.D:8372 ESTABLISHED
also i attached tcp connection between the nodes as above from source
and target,
can any one help why i have this behavior and how can i fix the delay,
thinking doing QoS or clean up and remoteexcution at that time ...

Thanks


More information about the cisco-nsp mailing list