cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
Announcements
Walkthrough Wednesdays
2869
Views
19
Helpful
19
Replies
Dennis Mink
Advisor

Video region/locations bandwidth allocation CUCM

I am struggling to understand the following,  (CUCM10.0.x. ) 

 

I have 2 phones, each in their own location and region, say A and B.  using G729 and 384kbps Video as the settings between the two regions.  When I make the call and check the available video bandwidth (for instance through RTMT performance monitoring ), I see that 500 kbps is deducted from the Location bandwidth to make that single video call happen. 

 

The 10.x SRND clearly states:

 

For example, assume that the link configuration for the location Branch 1 to Hub_None allocates 256 kbps of available audio bandwidth and 384 kbps of available video bandwidth. In this case the path from Branch 1 to Hub_None can support up to three G.711 audio calls (at 80 kbps per call) or ten G.729 audio calls (at 24 kbps per call), or any combination of both that does not exceed 256 kbps. The link between locations can also support different numbers of video calls depending on the video and audio codecs being used (for example, one video call requesting 384 kbps of bandwidth or three video calls with each requesting 128 kbps of bandwidth).

 

Well that is not what I am seeing, in fact, when I bring the location's video bandwidth to 499 kbps, I get no video 

 

anyone seeing the same, and what mechanism am I overlooking here?

 

 

 

Please remember to rate useful posts, by clicking on the stars below.

19 REPLIES 19

Much appreciated getting back to me on this one Ayo. 

You are right the allocated bandwidth cannot exceed the region setting, but what is eventually deducted from the Location is not the same as the maximum bandwidth settings. I reversed engineered this and found that g729. anmd G711 take 102kbps and 384 video deducts 500 kbps from the reqion.  Now I have those figures i have it all working,

Ever since I passed my CCNP Voice ages ago, my world was flat and in accordance with: 

excerpt from the 7.1.2 SRND re. Call admission control:

Cisco Unified Communications Manager continues to admit new calls to a link as long as sufficient bandwidth is still available. Thus, if the link to the Austin location in the example has 160 kb/s of available bandwidth, that link can support two G.711 calls at 80 kb/s each, six G.723 or G.729 calls at 24 kb/s each, or five GSM calls at 29 kb/s each. If any additional calls try to exceed the bandwidth limit, the system rejects them, the calling party receives reorder tone, and a text message displays on the phone.

Please remember to rate useful posts, by clicking on the stars below.

We have our own priorities aokanlawon . :P 

I took a stab at the traces, and also tried some stuff in lab. I think it is better for you to contact TAC for a detailed explanation.

I couldn't infer much from the CCM SDL traces as they were provided from a single server only. For LBM, however the traces are there from all nodes.

LBM is responsible for allocating the bandwidths and performing weight calculations for enhanced CAC.

The request to save bandwidth is made by the CCM always.

In our call, initial bandwidth request is 4000 kbps on both sides, which is then adjusted to 500 kbps in the end. I am unable to see the calculation of 500 as this is done on node 3 (CCM) and not node 2.

During initial call setup, we see approx. 4000 kbps allocated for bw. This is initiated by the 8941 phone. I think there is a TIAS field in the INVITE SDP because in my 8945 phone, I see TIAS value of 2000,000 .

14216662.000 |09:34:02.969 |SdlSig-O |PolicyAndCACRegisterReq |NA RemoteSignal |ReservationMgr(3,100,110,1) |SIPCdpc(2,100,82,134128) |3,100,14,30519.64^10.61.4.111^* |[R:N-H:0,N:1,L:0,V:0,Z:0,D:0] CI= 52418193 Branch= 0 reg=AU_VIC_Epping_REG cap=0 loc=47 MRGLPkid=7458a52f-3c9e-0920-63e0-ea722edee4e8 PrecLev=5 bearerlessCall=F VCall=T MMCap=0xb VCapCount=0 regiState=0 medReq=0 IsEmccD=F EmccDName=SEP5057A887BFDD rcId= ipMode=0 eoType=0 getPort=F sRTP=F cryptocap=0 tm=16 DTMF(wantRecep=0 provOOB=0 suppMeth=0 Cfg=1 PT=0 reqMed=1) honorCodec=F distMed=F mediaEP=F rsvpQoSType=0 qosFallback=F status=0 sipOfferNeededInd=T hasSDP=FLatentCaps=null videoTrafficClass=2 nonPreemptable=F geolocInfo={geolocPkid=, filterPkid=, geolocVal=, devType=4} CAL={v=-1, m=-1, tDev=F, res=F, devType=0} locPkid=3fa2293e-c8d0-eb59-8719-ab8a16c1bd80 locName=AU_VIC_Epping_LOC deductBW=F fateShareId=StandAloneCluster:52418193 videoTrafficClass=Desktop_TCL lbmCACInfo: retryVideoAsAudio=T overrideRegionBW=F sideACmDeviceType=4 isNewPartyRegWithLBM=F videoReservedKbps=3968 cacSpecificInfo.cacPartyState=0

The ReservationMgr process for this call runs on node 3, and we have node 2's traces for CCM. I needed to see mainly what ReservationMgr tells LBMInterface on Node 3.

We can thus see on LBM's traces that request for 4000 kbps initially comes in and allocated. This is just initial stages, so everything should be fine here still.

02451562.001 |09:34:02.988 |AppInfo |LBMServer - Received from CallManager:
msg = {
messageId = (101749)
msgkey = ()
nofwd = (0)
choice = (reservationReq)
reservationReq = (
transId = (9539942707464544135)
seqNum = (0)
choice = (reserve)
reserve = (
sideA = (
loc = (AU_NSW_Prestons_TP_LOC)
fs_id = (StandAloneCluster:52418192)
ext = ()
)
sideB = (
loc = (AU_VIC_Epping_LOC)
fs_id = (StandAloneCluster:52418193)
ext = ()
)
cacSt = (0)
precedLvl = (5)
mlpp = (0)
execOVR = (0)
enforce = (0)
a_val = (0)
v_val = (1)
i_val = (0)
a_bw = (0)
v_bw = (4000)
i_bw = (0)
a_op = (0)
v_op = (0)
i_op = (0)
ext = (<msg></msg>)
)
ext = ()
)
}

Now after call is completed, there is an adjust due to which this gets reduced to 500 kbps. Again I am not able to see what reservation mgr is doing here. But here is the request on LBM side:

02451603.001 |09:34:08.033 |AppInfo |LBMServer - Received from CallManager:
msg = {
messageId = (101756)
msgkey = ()
nofwd = (0)
choice = (reservationReq)
reservationReq = (
transId = (9539942707464544135)
seqNum = (0)
choice = (adjust)
adjust = (
sideA = (
loc = (AU_NSW_Prestons_TP_LOC)
fs_id = (StandAloneCluster:52418192)
ext = ()
)
sideB = (
loc = (AU_VIC_Epping_LOC)
fs_id = (StandAloneCluster:52418193)
ext = ()
)
cacSt = (1)
precedLvl = (5)
mlpp = (0)
execOVR = (0)
enforce = (0)
a_val = (0)
v_val = (1)
i_val = (0)
a_bw = (0)
v_bw = (500)
i_bw = (0)
a_op = (2)
v_op = (0)
i_op = (0)
ext = (<msg></msg>)
)
ext = ()
)
}

I can take a look at the traces again if you provide them from all nodes in cluster. Just select CCM and LBM traces for all the servers in cluster. I can take a look again.

Config snippets would help too!

Or you may contact TAC for a better answer.

Location and Regions are both different calculations. Both are not inter related.

~Avinash

I can't see how it is, the 10.x SRND has no mention of it. whereas previsous SRNDs mentioned location deductions. 

Please remember to rate useful posts, by clicking on the stars below.

Dennis,

 

Provide the traces from CCM as requested. Let's isolate and see what is happening on CUCM.

~Avinash

avinsrid89
Beginner

I don't think this problem should we observed.

Can you provide details CCM Service and Location Bandwidth Manager service traces from all nodes in cluster using RTMT. Set the detailed level of traces for all servers in cluster using this link as reference:

https://supportforums.cisco.com/document/126666/collecting-cucm-traces-cucm-862-tac-sr

~Avinash

Content for Community-Ad

Spotlight Awards 2021