[cisco-voip] WAN failure in call manager
Wes Sisk
wsisk at cisco.com
Fri Dec 11 09:39:50 EST 2009
yep! there are some interesting scenarios that arise where phones may
failover rather aggressively (geometric tcp). in recent cm versions this
is configurable on the phone configuration page. there are also
scenarios where router may artificially failover (dsp issues cause ios
to stop responding to crcx/dlcx/etc.)
I like your idea lelio. allows you to transition all devices at a
remote site as mostly a single entity. However, it depends on choosing
the appropriate trigger event. The trigger even may differ depending on
network topology and implementation
/wes
On Friday, December 11, 2009 9:21:28 AM, Lelio Fulgenzi
<lelio at uoguelph.ca> wrote:
> neat suggestion. i just read up on eem. pretty powerful stuff.
>
> on the corollary, you could also monitor when a phone registered
> before your MGCP gateways and then force the MGCP gateways to register
> as well using similar ACLs.
>
> i likes it!
>
>
> ---
> Lelio Fulgenzi, B.A.
> Senior Analyst (CCS) * University of Guelph * Guelph, Ontario N1G 2W1
> (519) 824-4120 x56354 (519) 767-1060 FAX (JNHN)
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> "Bad grammar makes me [sic]" - Tshirt
>
>
> ----- Original Message -----
> From: "Wes Sisk" <wsisk at cisco.com>
> To: "Lelio Fulgenzi" <lelio at uoguelph.ca>
> Cc: cisco-voip at puck.nether.net, "anand" <anand.eee at gmail.com>, "Karen
> Cheng" <kaz.cheng at gmail.com>
> Sent: Friday, December 11, 2009 9:13:12 AM GMT -05:00 US/Canada Eastern
> Subject: Re: [cisco-voip] WAN failure in call manager
>
> to be creative one could write an EEM script to monitor for gateway
> losing connectivity to mgcp call agent and then implement an ACL
> blocking all ip connectivity from the phone subnet toward the CM
> servers. that would effectively have the router control the phone
> rehome as well.
>
> /wes
>
> On Thursday, December 10, 2009 5:54:30 PM, Lelio Fulgenzi
> <lelio at uoguelph.ca> wrote:
>
> Good point Wes. That's what I was trying to say with respect to
> the last paragraph, but reading it again, I didn't make that point
> clear.
>
> Yes, the gateway does maintain connectivity for it's local ports
> (PRI, FXS, etc) and when it looses connectivity it triggers it's
> local ports to register via SRST to itself (or another SRST router).
>
> However, regardless of that, even if a router can still talk to
> the CUCM cluster, it will accept SRST registrations from phones.
> (see last message sent).
>
> But I love how you manage to start with "it depends.... " ;)
>
> ---
> Lelio Fulgenzi, B.A.
> Senior Analyst (CCS) * University of Guelph * Guelph, Ontario N1G 2W1
> (519) 824-4120 x56354 (519) 767-1060 FAX (JNHN)
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> "Bad grammar makes me [sic]" - Tshirt
>
>
> ----- Original Message -----
> From: "Wes Sisk" <wsisk at cisco.com>
> To: "Karen Cheng" <kaz.cheng at gmail.com>
> Cc: "Lelio Fulgenzi" <lelio at uoguelph.ca>,
> cisco-voip at puck.nether.net, "anand" <anand.eee at gmail.com>
> Sent: Thursday, December 10, 2009 5:45:03 PM GMT -05:00 US/Canada
> Eastern
> Subject: Re: [cisco-voip] WAN failure in call manager
>
> it depends.
>
> For MGCP functionality the gateway is monitoring CM. When failure
> is detected router follows the configured fallback apllication,
> usually h.323
> http://www.cisco.com/en/US/docs/voice_ip_comm/cusrst/admin/srst/configuration/guide/srs_wlan.html#wp1351962
>
> That is only for the gateway though.
>
> Phones monitor CM connecvity. When phone loses connectivity to CM
> they attempt to register with SRST router. When they register
> they tell the router about their configuration.
>
> To that end it is both phone initiated and gateway initiated. The
> gateway does not control all logic, nor does the phone control all
> logic.
>
> /Wes
>
> On Thursday, December 10, 2009 5:32:16 PM, Karen Cheng
> <kaz.cheng at gmail.com> wrote:
>
> Hi Lelio,
>
> I don't believe that part about the phone triggering the SRST
> function is true. From memory I believe the SRST router is the
> one that detects the WAN failure in which after a certain
> amount of time (dependent on how many servers in the CM
> cluster) it reconfigures itself to be the call agent instead.
> It then also continues to monitor connectivity to the CM
> cluster and once backup will again wait (dependent on how it
> is configured, wait until no calls are active etc) before
> handing the call agent functions back to the CM cluster.
>
> I may be wrong though.
>
> Karen
>
> On Fri, Dec 11, 2009 at 8:21 AM, Lelio Fulgenzi
> <lelio at uoguelph.ca <mailto:lelio at uoguelph.ca>> wrote:
>
> Correct, during the SRST mode, the CUCM cluster is not
> controlling anything. The phones register to the SRST
> router using a "simplified" configuration they store in
> memory. The tell the router what DNs it is supposed to
> have, and the router configures the appropriate
> dial-peers. Then using the dynamic dial peers it can call
> other phones, or using the static dial-peers (ones you
> program) a phone can call off-net using PSTN.
>
> Interesting thing is this, SRST is "always on" waiting for
> a device to register to it. The device is what controls
> when it talks to the SRST router or not. That's how you
> can test registering devices to the router when the WAN
> really isn't down.
>
> It also means that if there is some way or scenario that
> your phones may not be able to talk to CUCM, but your
> router can, you will have phones that can call each other,
> but not out the PSTN because the gateways are still
> registered to CUCM. That is, of course, if you have MGCP
> gateways. If you have H323 gateways, I think it's a moot
> point.
>
> Of course, I could be wrong on one or two points, and
> would be happy to learn something new. ;)
>
>
>
> ---
> Lelio Fulgenzi, B.A.
> Senior Analyst (CCS) * University of Guelph * Guelph,
> Ontario N1G 2W1
> (519) 824-4120 x56354 (519) 767-1060 FAX (JNHN)
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> "Bad grammar makes me [sic]" - Tshirt
>
>
>
> ----- Original Message -----
> From: "anand" <anand.eee at gmail.com
> <mailto:anand.eee at gmail.com>>
> To: cisco-voip at puck.nether.net
> <mailto:cisco-voip at puck.nether.net>
> Sent: Thursday, December 10, 2009 5:15:00 PM GMT -05:00
> US/Canada Eastern
> Subject: Re: [cisco-voip] WAN failure in call manager
>
> Hi,
>
> I have been breaking my head to understand the exact
> functions of SRST. During the WAN failure, SRST router
> takes PSTN path.My question is during this time will the
> call manager controls the call or Call Manager is
> completely out of operation(I mean it does not take any
> part in the call control on the PSTN path).
>
> _______________________________________________ cisco-voip
> mailing list cisco-voip at puck.nether.net
> <mailto:cisco-voip at puck.nether.net>
> https://puck.nether.net/mailman/listinfo/cisco-voip
>
> _______________________________________________
> cisco-voip mailing list
> cisco-voip at puck.nether.net <mailto:cisco-voip at puck.nether.net>
> https://puck.nether.net/mailman/listinfo/cisco-voip
>
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> cisco-voip mailing list
> cisco-voip at puck.nether.net
> https://puck.nether.net/mailman/listinfo/cisco-voip
>
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://puck.nether.net/pipermail/cisco-voip/attachments/20091211/3daa6f15/attachment.html>
More information about the cisco-voip
mailing list