On Fri, 2007-03-02 at 10:38 -0700, Scott Morris wrote: > We have a production server that is exhibiting some strange behavior. > We'll call this machine SERVER A.
This is why smart companies hire smart sysadmins instead of trained,
certified monkeys. Any number of things could be causing your problems.
It's going to help a lot if you understand the basics of the entire
architecture. Hardware and OS monitoring and tuning, network monitoring
and tuning, perhaps even MySQL monitoring and tuning. (After all, you
might have multiple problems. An OS problem, a network problem, and a DB
problem.)
As always, your first step should be to watch your logs as you test the
various connections. Next, you'll probably want to use a network sniffer
(Wireshark, AKA Ethereal) to watch what's happening.
It sounds like a problem I saw where SSH was default to IPV6 AAAA
lookups which the DNS server ignored. Eventually SSH fell back to IPV4 A
lookups and everything went full speed from then on. Of course, the slow
down was very predictable and your description makes it sound more
random.
On the other hand, it could be the server. Is it heavily loaded? Widely
varying times can be a result of concurrent load, especially heavy IO
and/or swapping.
Unfortunately, those are just a couple of random tips. Until you've
collected more info it's hard to be certain what's happening.
--
Stuart Jansen e-mail/jabber: [EMAIL PROTECTED]
google talk: [EMAIL PROTECTED]
"However beautiful the strategy, you should occasionally look at
the results." -- Winston Churchill
signature.asc
Description: This is a digitally signed message part
/* PLUG: http://plug.org, #utah on irc.freenode.net Unsubscribe: http://plug.org/mailman/options/plug Don't fear the penguin. */
