10-28-2008 08:49 AM
Hello,
I have this configuration:
serverfarm host SF_BOOKING
failaction purge
predictor leastconns slowstart 60
probe PROBE_HTTP_BOOKING
retcode 200 500 check count
rserver SERVER-1 80
conn-limit max 1200 min 1100
inservice
rserver SERVER-2 80
conn-limit max 1200 min 1100
inservice
rserver SERVER-3 80
conn-limit max 1200 min 1100
inservice
and state of servers is:
SKY-ACE1/www# sh serverfarm SF_BOOKING detail
serverfarm : SF_BOOKING, type: HOST
total rservers : 3
active rservers: 2
description : -
state : ACTIVE
predictor : LEASTCONNS
slowstart : 60 secs
failaction : PURGE
back-inservice : 0
partial-threshold : 0
num times failover : 0
num times back inservice : 0
total conn-dropcount : 0
Probe(s) :
PROBE_HTTP_BOOKING, type = HTTP
---------------------------------
----------connections-----------
real weight state current total failures
---+---------------------+------+------------+----------+----------+---------
rserver: SERVER-1
10.218.71.30:80 8 MAXCONNS 636 7530 140
max-conns : 1200 , out-of-rotation count : 0
min-conns : 1100
conn-rate-limit : - , out-of-rotation count : -
bandwidth-rate-limit : - , out-of-rotation count : -
retcode out-of-rotation count : 0
load value : 0
rserver: SERVER-2
10.218.71.35:80 8 OPERATIONAL 562 7297 127
max-conns : 1200 , out-of-rotation count : 0
min-conns : 1100
conn-rate-limit : - , out-of-rotation count : -
bandwidth-rate-limit : - , out-of-rotation count : -
retcode out-of-rotation count : 0
load value : 0
rserver: SERVER-3
10.218.71.40:80 8 OPERATIONAL 529 4712 58
max-conns : - , out-of-rotation count : -
min-conns : -
conn-rate-limit : - , out-of-rotation count : -
bandwidth-rate-limit : - , out-of-rotation count : -
retcode out-of-rotation count : 0
load value : 0
I have two questions:
1. Why is state of SERVER-1 'MAXCONNS', when real conns is under 1100 (min-conns)?
2. I find that SERVER-1 goes down (reason probe), client got TCP RST. Client is requesting VIP of course. When I clear 'conn-limit' to default (none), this problem isn't. The same if SERVER-3 is in this state (MAXCONNS). MAXCONNS is cancelled after 'no inservice, inservice'.
It's feature or bug? :)
sw release: 3.0(0)A1(8.0)
(I found no bug for this release)
thanks for any help,
martin
Solved! Go to Solution.
11-05-2008 01:43 AM
CSCsr20530: rservers are stuck in MaxConn state
CSCsr65774: server stuck in MAXCONN if conn-limit and RL hit simultaneously
CSCsq83367: server stuck in MAXCONN state if inservice/noinservice
CSCsr63601: rserver are stuck in MaxConn state after ft switchover
CSCsv21495: rservers stuck in Max-conn while configuring limits during traffic .
CSCsr72390: rserver do not get conn after going OOR w/ maxconn
11-04-2008 06:30 AM
Martin,
the maxconn limit function is broken in your release.
You should try A3(1) and see if it helps as many fixes where added already...but unfortunately there are still a few being worked on.
Gilles.
11-05-2008 01:07 AM
Gilles,
thank you for this information. It's very helpful for me, because I can't find it in release notes, or bug tool.
can you tell me, where can I find this information? (some release note, or bug ID).
thanks.
martin
11-05-2008 01:43 AM
CSCsr20530: rservers are stuck in MaxConn state
CSCsr65774: server stuck in MAXCONN if conn-limit and RL hit simultaneously
CSCsq83367: server stuck in MAXCONN state if inservice/noinservice
CSCsr63601: rserver are stuck in MaxConn state after ft switchover
CSCsv21495: rservers stuck in Max-conn while configuring limits during traffic .
CSCsr72390: rserver do not get conn after going OOR w/ maxconn
Find answers to your questions by entering keywords or phrases in the Search bar above. New here? Use these resources to familiarize yourself with the community: