Recently we have installed MySQL on our database cluster
which runs TurboLinux. The intention with the cluster is
to have one MySQL up at any given time. Some of the people
in our group proposed the following combination to check
on the status of MySQL server:
1. To look for the MySQL server process -- if there is none
running, then the server needs is automatically restarted.
2. A simple cron job performing write/select/delete with 2 records
on a test table in the test database is used to provide a second
layer of assureance. If task can not be performed, the MySQL
server is assumed having serious problems and is automatically
shutdown and restart afterwards.
My questions are the following.
1. Can anyone tell us what does "mysqladmin ping" provide us
on the status of MySQL server? It tells if the server is "alive".
Can someone explain what the "alive" mean?
2. What is a better way to provide gauges on the health of MySQL
For example we may be monitoring the number of locked up process
using "mysqladmin process". If there are tones of locked-up process
on 1 table, things may be wrong. If a process has locked up 1 table
for hours, we may need to look into the process. Could someone
provide us any suggestions?
Thanks very much.
831 440 5718
TAPESTRY.NET - On Target, Online Recruiting
100 Enterprise Way, Suite A2, Scotts Valley, CA 95066
Visit our web sites at:
http://www.asia-net.com Bilingual Professionals
http://www.developers.net Software Developers
http://www.itinfo.net IT Professionals