TMS unavailability

[brian.k.clarke]

Simple question to help a customer determine the value/worth of deploying an HA TMS (either via load balancer or active/standby)...

What functionality would be lost in the timeframe that TMS became unavailable for any reason?  Let's assume that all of the infrastructure and endpoint devices have been added to the system as managed devices, Movi/Jabber clients have been provisioned, phone books have been published and distributed and conferences scheduled. In other words, the system is fully configured/stable, and TMS goes down. In that timeframe prior to reviving/rebuilding it:

- Would there be any effect on existing managed devices, or would they continue to function 100% with their current configurations and phonebooks?

- What would be the effect on scheduled conferences? I'm thinking this might depend on the type of conference, in that an "Automatic Connect" (for example) requires TMS to kick the conference call off from the endpoint, but that maybe rendezvous or OBTP conferences would still function as normal w/o TMS being available - need confirm on all of this.

Thank you!

[Magnus Ohm]

Hi

Units will still operate but scheduled conferences set to start during the downtime will be affected. Phonebooks will be unavailable (corporate phonebooks) until TMS is back up again. CDR's produced and system messages to tms will be lost as tms is not there to collect them, however that is not a major issue in most cases. Systems will still be able to call and operate normally provisioned devices as well.

/Magnus

Sent from Cisco Technical Support iPad App

[brian.k.clarke]

Thanks for the response, Magnus - a couple follow-ups if I may (again, just to make sure I compeltely understand):

- Corporate phonebooks (whenever updated) aren't pushed out to the endpoints (initially and when updated on TMS)?  Just confirming that there's no way around having the corporate PB being unvailable if TMS is currently down.

- Could I get some additional clarification on how scheduled conferences will be "affected" (if scheduled to start during the outage window)? All conferences (including rendezvous, OBTP on endpoints, etc.), and/or just the "Automatic Connect" sessions?  Aren't most of these conferences (whether scheduled via TMS, Outlook TMS plug-in, etc.) just front-ends for the MCU?  If a standard conference bridge is scheduled successfully through TMS, then why wouldn't it be available for people to connect to even if TMS wasn't currently available?

Thank you!  A lot of questions here, I know, but there are a lot of conferencing options...

[Magnus Ohm]

Thanks for the response, Magnus - a couple follow-ups if I may (again, just to make sure I compeltely understand):

- Corporate phonebooks (whenever updated) aren't pushed out to the endpoints (initially and when updated on TMS)? Just confirming that there's no way around having the corporate PB being unvailable if TMS is currently down.

C-series and newer endpoints does direct lookups and cannot store the phonebook so there is no workaround for these endpoint(100phonebook contacts can be stored locally but this is a manual job pr endpoint). MXP series will store 400 entries in the phonebook if the tms goes down assuming coporate and downloaded (global phonebooks) are enabled in TMS.

- Could I get some additional clarification on how scheduled conferences will be "affected" (if scheduled to start during the outage window)? All conferences (including rendezvous, OBTP on endpoints, etc.), and/or just the "Automatic Connect" sessions? Aren't most of these conferences (whether scheduled via TMS, Outlook TMS plug-in, etc.) just front-ends for the MCU? If a standard conference bridge is scheduled successfully through TMS, then why wouldn't it be available for people to connect to even if TMS wasn't currently available?

The automatic connect wil definately not start. Not 100% sure about the obtp as tms usually creates the conference right before start so my guess is that the obtp conference will be affected as well (someone correct me if i'm wrong on this.. Dale??). The exchange integration is using the tms booking api and a conference created in outlook or tms does not make a difference. The last question is because the conference does not exist on the mcu until tms creates it upon start it does not create it upon booking if you follow.. A workaround would be to manually setup the conference on the mcu while tms is down if there is such a conference to start during downtime.

Having a redundant setup (hot spare tms) reduces the risk of critical downtime.

Hope this helps :)


Sent from Cisco Technical Support iPad App

[brian.k.clarke]

Thanks again for the feedback - I've reached out to the local Cisco team for their input on this as well, and will repost when I get further clarification.

[Kjetil Ree]

Hi,

[...] my guess is that the obtp conference will be affected as well (someone correct me if i'm wrong on this.. Dale??) [...]

TMS pushes OBTP items to endpoints a minute or two after the conference has been scheduled.

OBTP meetings are cleared if an endpoint reboots, but TMS resends them whenever it gets a boot event from an endpoint. But if TMS is down, it obviously cannot resend the OBTP meetings.

Regards,

Kjetil (not Dale )

[brian.k.clarke]

Thanks for the reply, Kjetil - one more follow-up, just to cover all conferencing options:

The Automatic Connect and OBTP explanations make 100% sense. I'm curious how the other TMS options (Manual Connect, No Connect and Reservation) might be impacted by a TMS outage. Reading through the TMS Administrator Guide, it sounds as if TMS is not directly involved in kicking off these types of conferences. For instance, if you create a "Manual Connect" conference, the "VC-Master" will get prompted to initiate the call (vs. TMS automatically doing so).

Would these other conference types function the same as OBTP?  Specifically, if I schedule a MC conference, is this information pushed down to the endpoint where it's RAM resident unless there's an endpoint restart (and thus, not affected by TMS outage)?

Thanks again

[Kjetil Ree · Accepted solution]

All multipoint conference types will be affected by TMS going down. This is because TMS doesn't do a "conference.create" command on the MCU until the conference is about to start.

"Manual connect": The prompt to the VC master to start the conference is sent by TMS, so you'll lose this conference type as well, just like OBTP and "automatic connect".

"Reservation" doesn't route the call and reserve ports and gateways, so such conferences should be unaffected by TMS becoming unavailable.

Specifically, if I schedule a MC conference, is this information pushed down to the endpoint where it's RAM resident unless there's an endpoint restart (and thus, not affected by TMS outage)?

It is dependent on the endpoint software version. Information is pushed to newer TC and TE endpoints so that they are visible in the "meetings" sections on the touch panels, but nothing is ever pushed to TC/TE on older software (pre 5.0, I believe) or to MXPs, Polycoms etc.

I'm not sure why the endpoint people have chosen to clear the list of meetings when the unit restarts. This is actually a problem for CUCM registered endpoints in TMS, where the boot event goes to the CUCM and not to TMS. We thus cannot guarantee that the meetings are resent after a boot.

[brian.k.clarke]

Thank you - very helpful!