Meetme Issue

[Yort Mantup]

Hello,

have a site that just recently starting having meetme conf issues where people are unable to join.  It's one of the features that I am not too familiar with but what I found is that in the Service Parameter configuration the Max MeetMe conf Unicast is set to 12 on our Pub and 13 on our Sub (1 Sub and 1 Pub Only).    Is that the sole determination on how many meetme conf. connections can occur in any one MeetMe conf?    I ran a CDR report showing the Conf Call Detail from that sites gateway.  The meeting with the issues showed 17 participants but I am told more were trying to join. (either received a busy signal or dropped call).  13 of those participants joined as an incoming call (dialed the meetme Direct Number) 4 were internal.  I did check that the sites PRI was not maxed out so that was not the issue.

With that said, how could 17 be connected at once?  Does meet me pool the two server resources allowing 25 connections? 

Appreciate the help,

[Jaime Valencia]

There's no way what you're telling can be right, that is a clusterwide parameter, so, you cannot have a different setting per server. Thus, no, there is no pooling resources as that is only one number.

There's no way you could have had 17 people if the parameter is set to 12. Only if you had advanced ad-hoc conferencing and conference daisy chaining would be possible.

That's the only parameter that sets how many users can join, if you need more than 12, adjust accordingly, FYI, I'm assuming your CFB does allow that many participants and the codecs you're using, if your CFB only allows 8 conferencees, that takes precedence as HW resources are finite.

[Jameson Gagnepain]

Are you using Software or Hardware conferencing resources? The phone that initiates the MeetMe conference will determine which conference resource to use based on that phone's Media Resource Group List configuration (MRGL is selected by the Device Pool if you don't override in the Phone configuration). It will use the same conference resource for the full duration of the call - there is no automatic joining of multiple conference resources, and any attempts to join the call past the limit imposed by the resource will result in a fast busy or a dropped call.

The "Maximum MeetMe Conference Unicast" Service Parameter is a clusterwide parameter... check it again, it should be the same for every server. If you are using Software conferencing resources, this parameter should limit total participants in the MeetMe to the number entered, including the conference initiator.

If you are using Hardware conferencing resources (for example, a PVDM3 on your voice gateway), you are limited by the "maximum conference-participants" setting within the dspfarm conference profile in the ios configuration.

I would suggest utilizing the Performance tab in RTMT to watch conference resources as callers join and leave the MeetMe if you're having trouble pinning down which resource is being used. Look for the "Cisco HW Conference Bridge Device" and "Cisco SW Conference Bridge Device" folders. The "ResourceActive", "ResourceAvailable", and "HWConferenceActive"/"SWConferenceActive" items are the most useful within those folders.

-Jameson

[Yort Mantup]

Thank you both for your responses.  Unsure what I was looking at or thinking since to your points, yes those are clusterwide paramaters and yes both are set to 12. 

I checked some of the phone configurations at that site and it appears that the phones MRG List setting is set to none.  I will need to figure out what the default setting is if set to none.  I am sure the site is using Software resources which according to your comments will limit each meet conference to a total of 12.  According to the report I ran, I do show a meetme conference with 17 attendees and am told they have had meeting in the past with more than 12 attendees.  I feel I am not getting the whole story from them yet.  They do have another call scheduled today at 3pm so I plan to utilize RTMT as Jameson suggested. 

[Jameson Gagnepain]

As I mentioned, if the conference initiator's phone does not have an MRGL set, it will utilize the MRGL defined in the Device Pool. If no MRGL is set for the Device Pool, then the "Default Media Resource Group" will be used... this is made up of all resources that are not in any Media Resource Group.

-Jameson

[Yort Mantup]

The device pool shows that the MRG List used is call "remotes".  The remotes has 3 selected MRG's (Base, RemoteMOH, XCODE in that order)

Is it correct that call will then use the Base resources first?  The base MRG is the group used by all devices.  It includes both ANN's, CFB's, XCODE and MTP's.

Plan to do some reading today to better understand but with the information above positive now that software resources are used.. with the config information given, it shouldn't change that the Service Parameter configuration the Max MeetMe conf Unicast is set to 12 and that is limit remains? 

[Jameson Gagnepain]

Yes, you are correct... the CFB resources in your "Base" MRG will be used first. If, for example, your "Base" MRG just has the CFB resources for your Pub and Sub in it:

1. MeetMe is started from a Cisco phone
2. Call Manager picks the CFB resource for either your Pub or Sub (picked by round-robin and available resources within each CFB)
3. That same CFB resource will be utilized until the MeetMe conference ends. Any limits on that resource will apply to the MeetMe conference - a SW CFB is limited to 128 total legs, and then your MeetMe parameter limit of 12 will limit participants as well.

So... say you only used SW resources with your 2 CUCM servers... you'd have a limit of 20 simultaneous 12-participant MeetMe conferences and 2 8-participant MeetMe conferences. Ad-hoc conferences would potentially reduce this total.

-Jameson

(Edited to correct references to XCODE/CFB)

[Yort Mantup]

 When I look at the XCODE MGR it shows 1 Transcoder which is a IOS Enhanced MTP followed by two MTP's (g711 and g729).  Would this then be using hardware resources on the router and not software? 

[Jameson Gagnepain]

The MRGs within your MRGL are used in order. The resources within each MRG are used in round-robin. So based on the order you gave earlier, the resources within the "XCODE" Media Resource Group would be used after everything in the "Base" and "RemoteMOH" Media Resource Groups. What transcoder resources are in your "Base" Media Resource Group?

-Jameson

[Yort Mantup]

correct, Base is the first listed.  There is one XCODE listed in the base MGR and that is the IOS Enhanced MTP (3825XCODE).   It also includes both ANN's, CFB's,  and MTP's.

ANNPubIP (ANN)

ANNSubIP (ANN)

CFBPubIP (CFB)

CFBSubIP (CFB)

GHNTEL3825XCODE (XCODE)

MTP_Pub (MTP)

MTP_Sub (MTP)

So if it is going to use the XCODE resources first based on your earlier scenario and the XCODE is using hardware resources then I would I need to look at the 3825 router and so a show dsp all to see how many resources are available.... sound correct or am I way off base?  I am following your logic 100%.. well done explaining and appreciated.  I first thought that the Meetme would be using SW resources but I believe it's utilizing the HW resources based off of the XCODE setup. 

[Jameson Gagnepain]

Oops! Just realized I mis-typed earlier... I meant to refer to the CFB resources, not XCODE. I'm editing my prior post to correct the error.

Given your list of resources in the Base MGR, I would expect "CFBPubIP" and "CFBSubIP" to be used first.

-Jameson

[Yort Mantup]

That did throw me off but figured before my prior post you meant CFB. :)  Ok, then Meetme's are using SW resources utilizing the CFB's first.  Which would then go back to the limit of 12 per Meetme conf is true.  I will monitor todays call and see what I see.  Thank again for you help.

[Yort Mantup]

as a follow up I used RTMT to monitor the conference.  It did use the Cisco SW Conf Bridge device.  I showed 1 active conference and a total of 10 resources active.  Just as you said Jameson  :)

Here's the question.  When I run a CDR Report (meetme only) for yesterday when the issues occurred. It shows there were two conferences where the no of participants joined was 14 and 17.  Each used the CFBPUB.  So how is this possible if the The "Maximum MeetMe Conference Unicast" Service Parameter is a clusterwide parameter is set at 12?    Is it possible that if a caller dials to the DID translating to that meet me number that it will instead use DSP resources and not SW?

[Jameson Gagnepain]

The resource used for a MeetMe is ALWAYS determined by the phone initiating the MeetMe. Once the conference begins, it will only use the resource that was allocated at the beginning of the call.

Are you sure the 14 and 17 are simultaneous participants? It seems more likely to me that you had some people who called, joined the conference, and at some point left the call and called back from another device. I see this behavior all the time in one of our weekly 8 am calls - people will call in from their cell phone on the way in, and then re-join the call from their desk phone when they arrive at work.

-Jameson
 

[CSCO12317494]

Jameson/Jaime,

Wanted to know the following regarding MeetMe conferencing:

1. I have several sites, using G729 between them and G711 within the same site. If a user initiates a call from site A and another from site B user joins the conference, is a transcoder used?
2. Also, if I have SW & HW CFBs configured for a device pool that the initiator's phone is allocated to, does CUCM combine these resources during an active conference?
3. How can I see the participants of a Meetme conference?