From: Chris Lewis \(chrlewis\) (chrlewis@cisco.com)
Date: Thu Sep 08 2005 - 22:06:28 GMT-3
Hi Godswill,
Thanks for your reply. My problem was not in the understanding of HSRP
operation, but in a generous, or biased reading of the dociumentation (I
guess trying a little too hard to find a way in which the documentation
could be correct). I have submitted a bug for this document, which
hopefully will result in it being corrected at some point.
I think you have a good understanding of HSRP, however there is one
point that needs correcting in your description.
" when R3 came back online even with its higher priority vlaue, it will
become the standy router. (this has nothing to do with tracking or the
interface serial you referred to or the prempting, this is the default
behaviour of hsrp peers)"
I think you meant to say it will NOT become the standby router. Here is
some more from the lab.
With the configuration in my initial e-mail (ie no preempt configured),
and R3 as the Active router, I reload R3 and get the following:
Router1#
*Sep 9 00:57:08.931: HSRP: Et0/0 Grp 1 Resign in 10.1.0.21 Active pri
105 vIP 10.1.0.1
*Sep 9 00:57:08.931: HSRP: Et0/0 Grp 1 Standby: i/Resign rcvd
(105/10.1.0.21)
*Sep 9 00:57:08.931: HSRP: Et0/0 Grp 1 Active router is local, was
10.1.0.21
*Sep 9 00:57:08.931: HSRP: Et0/0 Grp 1 Standby router is unknown, was
local
*Sep 9 00:57:08.931: HSRP: Et0/0 Grp 1 Standby -> Active
*Sep 9 00:57:08.931: %HSRP-6-STATECHANGE: Ethernet0/0 Grp 1 state
Standby -> Active
*Sep 9 00:57:08.931: HSRP: Et0/0 Grp 1 Redundancy "hsrp-Et0/0-1" state
Standby -> Active
*Sep 9 00:57:08.931: HSRP: Et0/0 Grp 1 Resign in 10.1.0.21 Init pri
105 vIP 10.1.0.1
*Sep 9 00:57:08.931: HSRP: Et0/0 Grp 1 Active: i/Resign rcvd
(105/10.1.0.21)
*Sep 9 00:57:08.931: HSRP: Et0/0 Grp 1 Coup out 10.1.0.22 Active pri
100 vIP 10.1.0.1
*Sep 9 00:57:11.971: HSRP: Et0/0 Grp 1 Redundancy group hsrp-Et0/0-1
state Active -> Active
*Sep 9 00:57:14.983: HSRP: Et0/0 Grp 1 Redundancy group hsrp-Et0/0-1
state Active -> Active
*Sep 9 00:57:55.783: HSRP: Et0/0 Grp 1 Standby router is 10.1.0.21
Router1#sho stand
Ethernet0/0 - Group 1
State is Active
2 state changes, last state change 00:04:27
Virtual IP address is 10.1.0.1
Active virtual MAC address is 0000.0c07.ac01
Local virtual MAC address is 0000.0c07.ac01 (bia)
Hello time 3 sec, hold time 10 sec
Next hello sent in 1.780 secs
Preemption disabled
Active router is local
Standby router is 10.1.0.21, priority 105 (expires in 7.400 sec)
Priority 100 (default 100)
Track object 100 state Up decrement 10
IP redundancy name is "hsrp-Et0/0-1" (default)
Router1#
R3 does not re-establish itself as the Active router as seen fomr the
following debugs, showing that it remains the standby router after the
re-boot.
R3#
*Sep 9 00:57:54.867: HSRP: Et0/0 Grp 1 Speak: d/Standby timer expired
(unknown)
*Sep 9 00:57:54.867: HSRP: Et0/0 Grp 1 Standby router is local
*Sep 9 00:57:54.867: HSRP: Et0/0 Grp 1 Speak -> Standby
*Sep 9 00:57:54.867: HSRP: Et0/0 Grp 1 Redundancy "hsrp-Et0/0-1" state
Speak -> Standby
Of course, if I put preempt in, I get the following debug output and R3
can re-establish itself as the Active router.
R3(config-if)#standby 1 preempt delay min 2
R3(config-if)#
*Sep 9 01:04:59.503: HSRP: Et0/0 Grp 1 Standby: h/Hello rcvd from lower
pri Active router (100/10.1.0.22)
*Sep 9 01:04:59.503: HSRP: Et0/0 Grp 1 Starting minimum preempt delay
(2 secs)
*Sep 9 01:05:02.555: HSRP: Et0/0 Grp 1 Minimum preempt delay expired
*Sep 9 01:05:02.555: HSRP: Et0/0 Grp 1 Active router is local, was
10.1.0.22
*Sep 9 01:05:02.555: HSRP: Et0/0 Grp 1 Standby router is unknown, was
local
*Sep 9 01:05:02.555: HSRP: Et0/0 Grp 1 Coup out 10.1.0.21 Standby pri
105 vIP 10.1.0.1
*Sep 9 01:05:02.555: HSRP: Et0/0 Grp 1 Standby -> Active
*Sep 9 01:05:02.555: %HSRP-6-STATECHANGE: Ethernet0/0 Grp 1 state
Standby -> Active
*Sep 9 01:05:02.555: HSRP: Et0/0 Grp 1 Redundancy "hsrp-Et0/0-1" state
Standby -> Active
*Sep 9 01:05:12.655: HSRP: Et0/0 Grp 1 Standby router is 10.1.0.22
R3(config-if)#do sho stand
Ethernet0/0 - Group 1
State is Active
2 state changes, last state change 00:00:45
Virtual IP address is 10.1.0.1
Active virtual MAC address is 0000.0c07.ac01
Local virtual MAC address is 0000.0c07.ac01 (bia)
Hello time 3 sec, hold time 10 sec
Next hello sent in 2.492 secs
Preemption enabled, delay min 2 secs
Active router is local
Standby router is 10.1.0.22, priority 100 (expires in 7.420 sec)
Priority 105 (configured 105)
Track object 100 state Up decrement 10
IP redundancy name is
-----Original Message-----
From: Godswill Oletu [mailto:oletu@inbox.lv]
Sent: Thursday, September 08, 2005 6:25 PM
To: Chris Lewis (chrlewis); Bob Nelson; ccielab@groupstudy.com
Subject: Re: Clarification of respoinse; HSRP and Track - 3550
Chris,
Though that work for you, but your interpretation of how HSRP, tracking,
etc might still need some clearification....
In hsrp, you can influence which router should be the active by
manipulating the priority value and just like dollars, more is better.
Once an active router has been elected, just like ospf DR/BDR, no
election takes place just because a router with a higher priority showed
up, with the same augument, if the priority of the active router is
reduced (eg by decreament value) even below the other standby router's
priority, the router is still going to be the active router its now
lower priority notwithstanding, your experiment below show that, despite
the fact that the priority of R3 was reduce to 95 which was less than
that of R1, R3 was still the active router. This is the default
behaviour of HSRP.
Each hsrp peer send keepalives to each other on their common interface
and if the active router do not response to keepalive after a predefined
period of time, the standy router will take over and become the new
active router.
This happened when you rebooted R3, R1 became the active router (you
will achieve the same result by shutting/no shutting the ethernet
interface of R3), when R3 came back online even with its higher priority
vlaue, it will become the standy router. (this has nothing to do with
tracking or the interface serial you referred to or the prempting, this
is the default behaviour of hsrp peers)
The prempt option changes this default behaviour. Prempt does not mean
that, the router with the prempt configured will always become the
active router.
Prempt just gives the router the power to call for or 'prempt-election'
each time it came online and like before the priority values of each
routers will determine who becomes the active router at the end of that
prempted election. The former active router will become the active again
if its priority is still the highest, however if during the downtime of
the active router with the prempt configure, if you increase the
priority of the standby router (now active ) above that of the down
active router, when the old active router with prempt comes online, it
will call for or prempt an election but it will fail become the active
router because their is another router with a higher priority.
Configuring prempt on both routers will only make sure that there is an
election each time one of the peers goes from up to down and back to up
state.
Configuring decreament on each or one of the routers without the prempt
option will only reduce its priority by the decreament amount if the
tracked interface goes down, but the routers will still retain their
various status (active or standby) because a new election was not
prempted. So tracking an interface/decreament go hand in hand with the
prempt option for it to work effectively.
You can test these behaviour in your lab and tell us what your
conclusions are. I have not tested these on a lab but that is my
understanding of how HSRP works.
HTH
Godswill Oletu
----- Original Message -----
From: "Chris Lewis (chrlewis)" <chrlewis@cisco.com>
To: "Bob Nelson" <nelsnjr@cox.net>; <ccielab@groupstudy.com>
Sent: Thursday, September 08, 2005 11:03 AM
Subject: Clarification of respoinse; HSRP and Track - 3550
> I'd like to clarify my previous post by examining the behavior of the
> commands in the command reference;
>
> I take the following configuration directly from the documentation and
> post it in to two routers in this configuration. I have to admit that
> upon closer reading of the documentation, I have to be extremely
> generous in my interpretation of the wording for it to be viewed as
> totally correct.
>
> S3/0
> |
> R3 (E0/0)----------(E0/0) R1 (e1/0)
>
> R3
> !
> track 100 interface serial3/0 ip routing
> !
> interface Ethernet0/0
> ip address 10.1.0.21 255.255.0.0
> standby 1 ip 10.1.0.1
> standby 1 priority 105
> standby 1 track 100 decrement 10
>
> R1
> track 100 interface E1/0 ip routing
> !
> interface Ethernet0/0
> ip address 10.1.0.22 255.255.0.0
> standby 1 ip 10.1.0.1
> standby 1 priority 100
> standby 1 track 100 decrement 10
>
> If I do a show standby on R1 immediately after entering these configs
I
> show this
>
> Router1#show standby
> Ethernet0/0 - Group 1
> State is Speak
> Virtual IP address is 10.1.0.1
> Active virtual MAC address is 0000.0c07.ac01
> Local virtual MAC address is 0000.0c07.ac01 (default)
> Hello time 3 sec, hold time 10 sec
> Next hello sent in 2.240 secs
> Preemption disabled
> Active router is 10.1.0.21, priority 105 (expires in 8.300 sec)
> Standby router is unknown
> Priority 100 (default 100)
> Track object 100 state Up decrement 10
> IP redundancy name is "hsrp-Et0/0-1" (default)
>
> Then after a while I see the following on R3
> R3#
> *Sep 8 14:27:49.191: %SYS-5-CONFIG_I: Configured from console by
> console
> *Sep 8 14:27:58.631: %HSRP-6-STATECHANGE: Ethernet0/0 Grp 1 state
> Standby -> Active
>
> If I check back on R1, I now see this
>
> Router1#show standby
> Ethernet0/0 - Group 1
> State is Standby
> 1 state change, last state change 00:00:23
> Virtual IP address is 10.1.0.1
> Active virtual MAC address is 0000.0c07.ac01
> Local virtual MAC address is 0000.0c07.ac01 (default)
> Hello time 3 sec, hold time 10 sec
> Next hello sent in 0.472 secs
> Preemption disabled
> Active router is 10.1.0.21, priority 105 (expires in 8.532 sec)
> Standby router is local
> Priority 100 (default 100)
> Track object 100 state Up decrement 10
> IP redundancy name is "hsrp-Et0/0-1" (default)
>
> Which shows R1 moving from Speak to Standby, R3 is now the active
> router. If I shut down S3/0 on R3, the priority is reduced to 95, but
it
> remains as the active router (but note the standby router has changed
to
> R1, which could be a trick wording in the exam), as shown by the
output
> of the debug standby terse and show standby command
>
> R3(config)#int s3/0
> R3(config-if)#shut
> R3(config-if)#
> *Sep 8 14:36:35.195: HSRP: Et0/0 Grp 1 Track 100 object changed,
state
> Up -> Down
> *Sep 8 14:36:35.195: HSRP: Et0/0 Grp 1 Priority 105 -> 95
> R3#
> *Sep 8 14:36:37.027: %LINK-5-CHANGED: Interface Serial3/0, changed
> state to administratively down
> *Sep 8 14:36:38.067: %LINEPROTO-5-UPDOWN: Line protocol on Interface
> Serial3/0, changed state to down
>
> R3#sho stand
> Ethernet0/0 - Group 1
> State is Active
> 2 state changes, last state change 00:17:56
> Virtual IP address is 10.1.0.1
> Active virtual MAC address is 0000.0c07.ac01
> Local virtual MAC address is 0000.0c07.ac01 (default)
> Hello time 3 sec, hold time 10 sec
> Next hello sent in 2.652 secs
> Preemption disabled
> Active router is local
> Standby router is 10.1.0.22, priority 100 (expires in 8.480 sec)
> Priority 95 (configured 105)
> Track object 100 state Down decrement 10
> IP redundancy name is "hsrp-Et0/0-1" (default)
>
> However if I reload R3, to simulate a failure of R3 (which is the way
I
> generously interpreted the phrase "if IP routing on serial interface
1/0
> in Router A fails") this happens:
>
> Router1#
> *Sep 8 14:47:28.631: HSRP: Et0/0 Grp 1 Resign in 10.1.0.21 Active
pri
> 95 vIP 10.1.0.1
> *Sep 8 14:47:28.631: HSRP: Et0/0 Grp 1 Standby: i/Resign rcvd
> (95/10.1.0.21)
> *Sep 8 14:47:28.631: HSRP: Et0/0 Grp 1 Active router is local, was
> 10.1.0.21
> *Sep 8 14:47:28.631: HSRP: Et0/0 Grp 1 Standby router is unknown, was
> local
> *Sep 8 14:47:28.631: HSRP: Et0/0 Grp 1 Standby -> Active
> *Sep 8 14:47:28.631: %HSRP-6-STATECHANGE: Ethernet0/0 Grp 1 state
> Standby -> Active
>
> I think I was too generous in my interpretation of what the
> documentation says. If it is relying on a failure of the router, there
> is no need to have the track serial command.
>
> Chris
>
> -----Original Message-----
> From: Chris Lewis (chrlewis)
> Sent: Thursday, September 08, 2005 8:29 AM
> To: 'Bob Nelson'; ccielab@groupstudy.com
> Subject: RE: HSRP and Track - 3550
>
> The quotes to be taken from the command reference and it does seem to
be
> correct. It does not talk about router A re-establishing itself after
it
> fails, which is what the preempt command does. I guess the example in
> the command reference is assuming the interfaces are brought up at the
> same time and they negotiate the router with higher priority as the
> active one. Regarding your second bullet, if Router A fails, Router B
is
> the only choice for taking over, regardless of priority or preempt.
>
> The discussion of preempt in the configuration guide below may be of
> additional help
>
>
http://www.cisco.com/univercd/cc/td/doc/product/software/ios122/122cgcr/
> fipr_c/ipcprt1/1cfip.htm)
>
> Chris
>
> -----Original Message-----
> From: nobody@groupstudy.com [mailto:nobody@groupstudy.com] On Behalf
Of
> Bob Nelson
> Sent: Wednesday, September 07, 2005 9:19 PM
> To: ccielab@groupstudy.com
> Subject: HSRP and Track - 3550
>
> Hello All:
>
> I want to make sure I am clear on the preempt and track for HSRP.
> Here is an example from the DocCD. My questions follow the example
>
> In the following example, the tracking process is configured to track
> the IP routing capability of serial interface 1/0. HSRP on Ethernet
> interface 0/0 then registers with the tracking process to be informed
of
> any changes to the IP routing state of serial interface 1/0. If the IP
> state on Serial interface 1/0 goes down, then the priority of the HSRP
> group is reduced by 10.
>
> If both serial interfaces are operational, then Router A will be the
> HSRP active router because it has the higher priority.
>
> However, if IP routing on serial interface 1/0 in Router A fails, then
> the HSRP group priority will be reduced and Router B will take over as
> the active router, thus maintaining a default virtual gateway service
to
> hosts on the 10.1.0.0 subnet.
> Router A Configuration
> track 100 interface serial1/0 ip routing !
> interface Ethernet0/0
> ip address 10.1.0.21 255.255.0.0
> standby 1 ip 10.1.0.1
> standby 1 priority 105
> standby 1 track 100 decrement 10
>
> Router B Configuration
> track 100 interface serial1/0 ip routing !
> interface Ethernet0/0
> ip address 10.1.0.22 255.255.0.0
> standby 1 ip 10.1.0.1
> standby 1 priority 100
> standby 1 track 100 decrement 10
>
> 1. "If both serial interfaces are operational, then Router A will be
the
> HSRP active router because it has the higher priority."
> Is this necessarily true. Does it not depend on which interface
> was brought up first?. Without preempt, A will not become active if
it
> is brought up second, correct?
> Without the preempt command on Router A, it will not take over
> just by virtue of the higher priority.
>
> 2. "if IP routing on serial interface 1/0 in Router A fails, then the
> HSRP group priority will be reduced and Router B will take over as the
> active router"
> Again, just because the priority on A (assuming it is the active)
> drops below router B's priority, does not mean it will take over as
> active,
> without B having preempt configured. Additionally, if B is
> configured
> to preempt A when Serial 1/0 goes down on router A, so it become the
> active,
> A must also have preempt configured to resume its active role
when
> the
> serial comes up.
>
> Is this correct and they just did not put the preempt statements for
> brevity(confuse me) or am I missing the concept.
>
> Thanks for the help. Sorry for the length, wanted to be complete
>
>
This archive was generated by hypermail 2.1.4 : Sun Oct 02 2005 - 14:40:14 GMT-3