|
|
 FAQ
 Search
 Memberlist
 Usergroups
 Register
 Profile
 Private messages
 Log in
|
|
| Author |
Message |
BernieM
Guest
|
 Posted: Tue Apr 18, 2006 1:45 pm Post subject: IP SLB on 6500 series |
 |
|
Hello all,
Having trouble with slb. A server keeps cycling from operational to failed
and back again ... over and over ...
Apr 18 19:22:48: %SLB-6-REAL: Real 172.22.2.125 (VPRDNS) has changed state
to FAILED
Apr 18 19:22:51: %SLB-6-REAL: Real 172.22.2.125 (VPRDNS) has changed state
to OPERATIONAL
Apr 18 19:22:56: %SLB-6-REAL: Real 172.22.2.125 (VPRDNS) has changed state
to FAILED
But the probe never shows any misses:
wcq15-core-6513#sh ip slb probe name vprdns detail
VPRDNS, dns, address = 0.0.0.0, interval = 15, faildetect = 3
VPRDNS, type = server
target = 172.22.247.1:53, real = 172.22.2.125:0, virtual =
172.22.247.1:53 TCP
state = OPERATIONAL, status = 1, operation id = 170 Server NAT
outages = 0, failures = 0, successes = 208, tests = 0
current = never, cumulative = 00:00:00
target = 172.22.247.1:53, real = 172.22.2.125:0, virtual =
172.22.247.1:53 UDP
state = OPERATIONAL, status = 1, operation id = 169 Server NAT
outages = 0, failures = 0, successes = 208, tests = 0
current = never, cumulative = 00:00:00
Here's the config:
ip slb vserver VPRDNS
virtual 172.22.247.1 udp dns
serverfarm VPRDNS
idle 5
replicate casa 172.22.31.253 172.22.31.252 1029
synguard 200
inservice standby VPRDNS-HSRP
!
ip slb serverfarm VPRDNS
nat server
nat client CLIENTS
probe VPRDNS
!
real 172.22.2.125
weight 1
inservice
!
real 172.22.9.50
weight 1
inservice
!
real 172.22.16.207
weight 1
inservice
!
ip slb probe VPRDNS dns
lookup prproxy.ourdomain.com
interval 15
Also, the VIP only forwards UDP but the probe is testing both UDP and TCP.
Can I control this?
Thanks,
BernieM |
|
| Back to top |
|
 |
|
|
BernieM
Guest
|
| Posted: Wed Apr 19, 2006 11:04 am Post subject: Re: IP SLB on 6500 series |
|
|
"BernieM" <c@fight.com> wrote in message
news:qz21g.9399$vy1.3962@news-server.bigpond.net.au...
| Quote: | Hello all,
Having trouble with slb. A server keeps cycling from operational to
failed and back again ... over and over ...
Apr 18 19:22:48: %SLB-6-REAL: Real 172.22.2.125 (VPRDNS) has changed state
to FAILED
Apr 18 19:22:51: %SLB-6-REAL: Real 172.22.2.125 (VPRDNS) has changed state
to OPERATIONAL
Apr 18 19:22:56: %SLB-6-REAL: Real 172.22.2.125 (VPRDNS) has changed state
to FAILED
But the probe never shows any misses:
wcq15-core-6513#sh ip slb probe name vprdns detail
VPRDNS, dns, address = 0.0.0.0, interval = 15, faildetect = 3
VPRDNS, type = server
target = 172.22.247.1:53, real = 172.22.2.125:0, virtual =
172.22.247.1:53 TCP
state = OPERATIONAL, status = 1, operation id = 170 Server NAT
outages = 0, failures = 0, successes = 208, tests = 0
current = never, cumulative = 00:00:00
target = 172.22.247.1:53, real = 172.22.2.125:0, virtual =
172.22.247.1:53 UDP
state = OPERATIONAL, status = 1, operation id = 169 Server NAT
outages = 0, failures = 0, successes = 208, tests = 0
current = never, cumulative = 00:00:00
Here's the config:
ip slb vserver VPRDNS
virtual 172.22.247.1 udp dns
serverfarm VPRDNS
idle 5
replicate casa 172.22.31.253 172.22.31.252 1029
synguard 200
inservice standby VPRDNS-HSRP
!
ip slb serverfarm VPRDNS
nat server
nat client CLIENTS
probe VPRDNS
!
real 172.22.2.125
weight 1
inservice
!
real 172.22.9.50
weight 1
inservice
!
real 172.22.16.207
weight 1
inservice
!
ip slb probe VPRDNS dns
lookup prproxy.ourdomain.com
interval 15
Also, the VIP only forwards UDP but the probe is testing both UDP and TCP.
Can I control this?
Thanks,
BernieM
|
Forget the question about the probe testing both UDP and TCP. We have
another (tcp) vip configured using the same server farm and probe so it's
expected.
I did discover today though that although the probe shows no misses, "show
ip slb real name vprdns detail" does show some 'connection failures' which
increment in parallel with the server becoming 'failed'. Ethereal tracing
shows all connection attempts successful though.
BernieM |
|
| Back to top |
|
|
|
|
