[cisco-voip] [External] Re: MRA failover doesn't work, Cisco TAC agrees, says it's a documentation defect

Tue Jun 21 13:04:06 EDT 2022

I'm a bit miffed on the need for the extra expressway C. We have very few MRA phones, but hadn't had this type of problem, an expressway is somehow busted and not accepting registrations - did they offer any explanation as to why that piece is needed?

The only thing I'd go to look up is how the CM list is being populated, if changing the CM group to bump the shut subscriber down (assuming reg order is sub -> pub), just since that'd come up before. Expressway doesn't seem to be configured to be aware that a UCM has gone away, despite the zone going down, at least for UDS and discovery.  I'm sure TAC looked at that though. I have a conversation going with them about this and Jabber SSO for a similar reason, that the device's configuration isn't dynamic to represent the state of the infrastructure, and sometimes they get stuck trying something that won't work and fail despite other components being available to serve them. That probably doesn't help with anything other than to say we're in a similar boat, just with Jabber and MRA.

Adam Pawlowski
Network Engineer | Network and Communication Services
University at Buffalo Information Technology (UBIT) 
243 Computing Center, Buffalo, NY 14260 

> -----Original Message-----
> From: cisco-voip <cisco-voip-bounces at puck.nether.net> On Behalf Of
> Matthew Huff
> Sent: Tuesday, June 21, 2022 12:54 PM
> To: Hunter Fuller <hf0002 at uah.edu>
> Cc: cisco-voip voyp list <cisco-voip at puck.nether.net>
> Subject: Re: [cisco-voip] [External] Re: MRA failover doesn't work, Cisco TAC
> agrees, says it's a documentation defect
> 
> We have no interest in setting up a jabber environment in order to debug
> ciscos's issue.
> 
> Yes, every expressway-e knows about all expressway-c, all expressway-c
> know about CUCM. Cisco TAC has verified the configuration, logs, and
> diagnostic. I've been working with them for 2 months and it's been escalated
> to backline-engineering. They looked at the Cisco Phone PRT logs and
> confirmed that it's a known limitation, and there is no solution.
> 
> Maybe it's an issue with later versions of CUCM and/or expressway? We are
> running the latest including latest phone firmware.
> 
> Failover works great except in one scenario where both the CUCM subscriber
> and the expressway-c that reside on the same machine are both shut down.
> Brining either one up, and the phone registers.
> 
> 
> -----Original Message-----
> From: Hunter Fuller <hf0002 at uah.edu>
> Sent: Tuesday, June 21, 2022 12:41 PM
> To: Matthew Huff <mhuff at ox.com>
> Cc: Kent Roberts <dvxkid at gmail.com>; cisco-voip voyp list <cisco-
> voip at puck.nether.net>
> Subject: Re: [External] Re: [cisco-voip] MRA failover doesn't work, Cisco TAC
> agrees, says it's a documentation defect
> 
> It might be worth setting up a Jabber test endpoint just to see.
> 
> Some questions though:
>  - Does every Expressway-E know about every Expressway-C?
>  - Does every Expressway-C know about every CUCM?
> 
> I'm trying to figure out what the desired architecture is, and/or how this
> problem would happen.
> In our environment, the above are both true. So the loss of any number of
> anything, should not result in failover issues - and that is the behavior we
> have seen (we have shut down entire sites due to maintenance, power
> failure, etc. and failover worked).
> In fact, we have found MRA phones to be great at failover in this way (our
> MRA phones are all 8851s). Jabber has been the problem child.
> 
> --
> Hunter Fuller (they)
> Router Jockey
> VBH M-1C
> +1 256 824 5331
> 
> Office of Information Technology
> The University of Alabama in Huntsville
> Network Engineering
> 
> On Tue, Jun 21, 2022 at 9:13 AM Matthew Huff <mhuff at ox.com> wrote:
> >
> > We don’t use Jabber nor Webex.
> >
> >
> >
> > Cisco TAC has been escalated and they have been working on this for over
> 2 months. I have sent repeated expressway and PRT logs from the phone.
> After working with Cisco engineering, the claim it is “working as intended”
> and plan on updating the documentation to reflect the limitation that if you
> loose both the subscriber and redundant expressway-C server, failover
> won’t happen.
> >
> >
> >
> > I’d love to be proven wrong since we may have to completely replace our
> solution.
> >
> >
> >
> >
> >
> > From: Kent Roberts <dvxkid at gmail.com>
> > Sent: Tuesday, June 21, 2022 10:09 AM
> > To: Matthew Huff <mhuff at ox.com>
> > Cc: cisco-voip voyp list <cisco-voip at puck.nether.net>
> > Subject: Re: [cisco-voip] MRA failover doesn't work, Cisco TAC agrees,
> > says it's a documentation defect
> >
> >
> >
> > This sound more like a config issue…
> >
> >
> >
> > Have run into issues where expressways go stupid when boxes go offline
> >
> > As for it being the phones 88xx. Does the same happen with jabber or
> webex?    If it does i’d  requeue the case….
> >
> >
> >
> > Kent
> >
> >
> >
> > On Jun 21, 2022, at 07:47, Matthew Huff <mhuff at ox.com> wrote:
> >
> > 
> >
> > We have a fairly common and standard deployment for our MRA solution.
> > All are running CUCM 14+, latest Expressway, etc…
> >
> >
> >
> > Vmware server 1 (jn DMZ)
> >
> >   ExpressWay-E-1
> >
> >
> >
> > Vmware server 2 (in DMZ)
> >
> >    ExpressWay-E-2
> >
> >
> >
> > Vmware Server 3 (In Core)
> >
> >   CUCM Publisher
> >
> >    Expressway-C-1
> >
> >
> >
> > VMWare Server 4( In Core)
> >
> >    CUCM Subscriber
> >
> >    Expressway-C-2
> >
> >
> >
> >
> >
> > If ether Expreway-E VMs fail, redundancy works fine If either CUCM
> > fails, redundancy works fine If either Expressway-C VMs fail,
> > redundancy works fine If VMWare Server 4 fails (say during patching,
> > hardware maintenance or hardware failure), redundancy fails. Remote
> phones un-register and never register no matter what is done. If either
> CUCM Subscriber or Expressway-C-2 is brought back online, phones register.
> >
> >
> >
> > Cisco TAC claims that this is a limitation of our Cisco 88xx SIP MRA phones
> and is not solvable unless we purchase two new vmware servers and split
> the CUCM and Expressway-C into separate servers so they both won’t go
> down at once. Sinc VMWare Server 3 & 4 are at different locations, vMotion
> isn’t an option since there is no shared storage.
> >
> >
> >
> > Anyone run into this or have any suggestions? We have engaged our VAR
> and cisco rep and may have to replace our phone system since we are all
> working from home and MRA support including redundancy is critical to us.
> >
> >
> >
> >
> >
> >
> >
> > _______________________________________________
> > cisco-voip mailing list
> > cisco-voip at puck.nether.net
> >
> https://nam12.safelinks.protection.outlook.com/?url=https%3A%2F%2Fpuck
> > .nether.net%2Fmailman%2Flistinfo%2Fcisco-
> voip&data=05%7C01%7Cajp26
> >
> %40buffalo.edu%7C719bda6c11134986ee9d08da53a6afaa%7C96464a8af8ed4
> 0b199
> >
> e25f6b50a20250%7C0%7C0%7C637914272634938090%7CUnknown%7CTWFp
> bGZsb3d8ey
> >
> JWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%
> 7C300
> >
> 0%7C%7C%7C&sdata=rBPHyezmLEzor0096y7a9yyEuMv83QEn2tRd9a%2
> BI%2FsU%3
> > D&reserved=0
> >
> > _______________________________________________
> > cisco-voip mailing list
> > cisco-voip at puck.nether.net
> >
> https://nam12.safelinks.protection.outlook.com/?url=https%3A%2F%2Fpuck
> > .nether.net%2Fmailman%2Flistinfo%2Fcisco-
> voip&data=05%7C01%7Cajp26
> >
> %40buffalo.edu%7C719bda6c11134986ee9d08da53a6afaa%7C96464a8af8ed4
> 0b199
> >
> e25f6b50a20250%7C0%7C0%7C637914272634938090%7CUnknown%7CTWFp
> bGZsb3d8ey
> >
> JWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%
> 7C300
> >
> 0%7C%7C%7C&sdata=rBPHyezmLEzor0096y7a9yyEuMv83QEn2tRd9a%2
> BI%2FsU%3
> > D&reserved=0
> _______________________________________________
> cisco-voip mailing list
> cisco-voip at puck.nether.net
> https://nam12.safelinks.protection.outlook.com/?url=https%3A%2F%2Fpuck
> .nether.net%2Fmailman%2Flistinfo%2Fcisco-
> voip&data=05%7C01%7Cajp26%40buffalo.edu%7C719bda6c11134986ee
> 9d08da53a6afaa%7C96464a8af8ed40b199e25f6b50a20250%7C0%7C0%7C6379
> 14272634938090%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiL
> CJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&a
> mp;sdata=rBPHyezmLEzor0096y7a9yyEuMv83QEn2tRd9a%2BI%2FsU%3D&a
> mp;reserved=0