[imapproxy] Re: [imp] serious mysql or imap problems on a imp
server
Didi Rieder
adrieder@sbox.tugraz.at
Fri Oct 25 13:07:22 2002
--On Thursday, October 24, 2002 01:29:05 PM -0500 Eric Rostetter <eric.rostetter@physics.utexas.edu> wrote:
> Quoting Didi Rieder <adrieder@sbox.tugraz.at>:
>
>> If other info is required please let me know.
>
> Well, for both the file I'm writing, and for debugging your problem, what
> are your settings in the imapproxy configuration file, and have you tried
> tweaking them? (Things like keepalive, client/server timeout, max_reuse,
> etc).
Here my settings:
remote_address sbox.tugraz.at
local_address 127.0.0.1
remote_port 143
local_port 1430
accept localhost
debug off
log_facility LOG_LOCAL6
logging on
keepalive 60
client_timeout 60
server_timeout 1800
stats_frequency 60
max_reuse 30
However, I have done some "simple" performance tests without IMP just with
basic IMAP connections and commands using the following expect script:
#!/usr/local/bin/expect -f
set timeout 30
# no proxy
#spawn telnet sbox.tugraz.at 143
# proxy
spawn telnet localhost 1430
# no proxy
# "* OK sbox.tugraz.at Cyrus IMAP4 v2.0.16 server ready" {
# proxy
# "* OK IMAP4rev1 ImapProxy v1.0-1 Welcome to my dominion." {
expect {
"* OK IMAP4rev1 ImapProxy v1.0-1 Welcome to my dominion." {
send "0 login user pass\r";
exp_continue
}
"0 OK" {
send "1 LIST * *\r";
exp_continue
}
"1 OK" {
send "2 SELECT INBOX\r";
exp_continue
}
"Completed" {
send "3 logout\r";
exp_continue
}
"3 OK"
{
exit 0;
}
}
exit 0;
I called the script in a loop to execute it 30 times and timed the duration.
Here the results:
direct:
real 0m28.907s
user 0m0.890s
sys 0m1.220s
Proxy:
real 0m20.911s
user 0m0.650s
sys 0m1.330s
This results show, that there is a significant performance gain when I use the
proxy. So I guess the problem must be related to IMP/c-client or wath erver,
since basic IMAP works great!
Any ideas?
Didi
p.s: sorry for suspecting the proxy :-)
--
-------------------------
Didi Rieder
adrieder@sbox.tugraz.at
-------------------------