We have a CSS switch running 4.01 (Build 50). We are load-balancing connections to dual servers behind the CSS. When the service to one of the servers goes down/suspended then returns to active we don't get any new connections to that service, unless we reboot the switch. Does anyone have any ideas on why this might be happening? Also, I can post the config/rules if anyone can help.
Here is the rule...
> Name: IBOFX_Rule Owner: IB
Author: Local Index: 1
State: Active Type: HTTP
IP Redundancy: Master
Header Field Group:
Total Bytes: 105134367 Total Frames: 234433
Total Redirects: 0 Total Rejects: 4
Overload Rejects: 0
Balance: Round Robin
Advanced Balance: None
Sticky Mask: 255.255.255.255
Sticky Inactivity timeout: 0 minutes
Sticky No Cookie Found Action: Balance
Sticky Server Down Failover: Balance
ArrowPoint Cookie Path: /
ArrowPoint Cookie Expiration: Browser Exit
ArrowPoint Cookie Text: None
ArrowPoint Cookie Url: None
Arrowpoint Cookie display url in header: Enabled
String Match Criteria:
String Range: 1 - 100
String Prefix: ""
String Eos-Char: "" String Ascii-Conversion: Enabled
String Skip-Len: 0 String Process-Len: 0
String Operation: Match-Service-Cookie
Local Load Threshold: 254
Name: Hits: Wgt: State: Ld: KAlive: Conn: DNS:
----- ----- ---- ------ --- ------- ----- ----
IBOFX1 46,461 S-1 Alive 2 SCRIPT 7 0
IBOFX2 48,607 S-1 Alive 2 SCRIPT 5 0
DNS Names: DNS TTL:
DNS Balance: roundrobin
Size: 10, Type: HitCount, Threshold 0, Interval 1
We are having the same issue, did you ever resolve it ?
What we notice is that if a request comes to a service before the switch notices that it is down and tries to send a request to it, it never gets another connection. I have to stop the service then start it again to resolve it.
I would love to hear if you have solved this or not.
We have not resolved the issue. We tried upgrading to a newer IOS, but it did not correct the problem. We are currently reviewing our configuration to make sure that we have all the rules setup properly.