- To: "<Unknown>MailList-SLUG" <slug@xxxxxxxxxxx>
- Subject: [SLUG] PostgreSQL and httpd weirdness
- From: Howard Lowndes <lannet@xxxxxxxxxxxxx>
- Date: Mon, 27 Sep 2004 15:13:30 +1000
- Organization: LANNet Computing Associates
I have two servers which interact with each other.
One is a web server, let's call it W and the other is a PostgreSQL
server, let's call it S.
W sends sql commands to S, which sends back the response.
W sends a 50 byte request to S with its seq number correctly set
(3622504195) and the ack number correctly set to match the last seq from
S plus the payload of the last packet from S (3658609267)
S responds with a 1448 byte payload, it's seq is correct (3658609267)
and the ack reflects correctly the 50 byte payload from W (3622504245)
S then does a push ack with a payload of 126 so it's seq rises to
3658610715 and the ack remains at 3622504245. This is the second part
of the response from S to W.
At this point W appears to lose the plot and instead of sending a zero
payload ack with seq 3622504245 and ack 3658610841 (= 3658610715 + 126),
it send a zero payload ack with seq 3622504245 and ack of 3658609267, as
if it never received the 1448 byte packet from S as well as ignoring the
126 byte push packet.
Needless to say S resend the big packet but W keeps insisting that it
wants it again.
At this point everything just hangs.
What would this be symptomatic of? MTU problems? NIC problems? or
what. I though it might be RAM but I have just recently upped that.
Both opsys are RH9.0.
--
Howard.
LANNet Computing Associates;
Your Linux people <http://www.lannetlinux.com>
------------------------------------------
"When you just want a system that works, you choose Linux;
when you want a system that just works, you choose Microsoft."
------------------------------------------
"Flatter government, not fatter government;
Get rid of the Australian states."