1.  Connect to the PUBLISHER web interface

 2.  OS Administration --> Settings à NTP Servers --> Add  new
    a.  IP address:  <ntp IP>
    b.  SAVE
    c.  Refresh page and wait to see NTP accessible
    d. Connect on CUCM CLI (ssh) and type: utils ntp status (see synchronised to NTP server..)

PS: Subscriber will always use the CUCM Publisher as ntp server.

Clock timezone CET +1
Clock summer-time CEST recurring first SUN April 01:00 last SUN Oc 01:00 < offset> (default 60)
ntp source loopback0
ntp server <ntp ip addr>
show ntp status (see synchronize, this may take up to 30 minutes)

To use the correct timezone for logs and debugs, you need to add this cmds:

For debug: service timestamps debug datetime msec localtime show-timezone year
For logs: service timestamps log datetime msec localtime show-timezone year

1.  What codec used?
2. cRTP?
3. What L2 technology used?
      
1.
 
G711 (20 sampling rate):160
G729 (20 sampling rate): 20

Other Sampling rate:

G711 30 samp rate: 240
G729 30 samp rate: 30

If Sample rate = 20 at 50pps
if sample rate = 30 at 33pps
      
2. cRTP

 If yes= 2 bytes header or 4 if checksum enabled (disabled
  by default but use 4 bytes..)
if NO= 40 byte
header


3.  L2 technology used:

Ethernet: 32
MLP: 13
Frame-relay:4 (+ 2) ! SRND does not add the 2 bytes for frame delimiter! à Frame-relay=6
Frame-relay with LFI FRF12: 8
PPP:12

Calculate:
Bandwidth=
((codec+header+L2)*8) * pps (50 or 33)
example:
G729
no cRTP
Frame relay
(use all default..)

Bandwidth = ((20+40+6)*8)*50 =
26400
25600/1000=26.4
Kbps per call

Enable the CallManager Advanced service parameter in CUCM:
“Allow Peer to Preserve H.323 Calls”  

on the Gateway:
voice service voip
   call preserve
   no h225 timeout keepalive

1. Start by reading all questions
2. Note all important points and routing requirments
3. Start configuration by device: Switch, R1, R2, R2, CUCM, CUC, CUPS, UCCX
4. Test your configuration (1 hour)
5. Verify, Verify, Verify (if u have time left)

Device-Based Strategy as explained by CCIE Voice GURU (http://ciscovoiceguru.com/1878/device-based-lab-strategy/)

Skinny phones = POTS dial peers with voice ports!!
SIP phones = voip dial peers (no voice port created by the SIP phone)


To configure SIP Phones on CME:

Voice register global
 mode cme ( by default it is SRST mode, to put it back: “no mode cme”)
 source-address <CME IP address> port 5060

 max-dn 2
 max-pool 2   ( = max-ephone )
 tftp-path flash: (the location where the config files will be stored, we can choose a subfolder ..)
 create profile ( = create cnf-files BUT we should put it after any change to create config file!! Not a 1 time command)

Voice register dn 1 (= ephone-dn 1)
 number 3005

 Voice register pool 1 (= ephone 1)
 id mac 001B.D4C6.1AA1
 type 7960
 number 1 dn 1 (= button 1:1)
 dtmf-relay rtp-nte
 description 3214-3005  (label to show on the phone top)
 codec g711ulaw

See config files:
Show voice register tftp-binding ( show SIP phones config files) SIPXXXXXXXXXXXX.cnf
Show telephony-service tftp-binding
( show Skinny phones config files) SEPXXXXXXXXXXXX.cnf.xml

 !!Without these files the phone will not be able to register

To debug SIP Phone registration:
debug tftp event
debug ccsip mess

Check to see these messages: 
1.      “REGISTER sip:10.10.202.1 SIP/2.0” message
2.      SIP/2.0 100 Trying
3.      SIP/2.0 200 OK

!if a SIP phone attempt to register, it will not register with “Voice register global”, we need to setup the voice registrar server:

Voice service voip
  registrar server expire max 600 min 60 ( the registration will expire in 600 minutes max and 60 min max)

!!Unlike SCCP Phones, if we have a dn on the phone this does not mean that the phone is registered with CME!! Because SIP is a “peer to peer” protocol.

Without the registration to CME (if we don’t configure registrar server), we will get dial tone and we will be able to dial another phone BUT inbound calls will not be possible. The CME will not route calls to the phone!!

To test if a SIP phone is registered:
The registration will be done when a 100 trying then a 200 ok message comes in.

Show voice register all
show voice register dial-peers ( see the voip dial peer built point to the IP address of the phone)
show voice register pool 1 (see phone + created dial peer)
+ test an inbound call to that phone.

Dial peer for phone:

Dial-peer voice 40001 voip
  destination-pattern 3005
  session target
  ipv4 :10.10.202.51 :5060
  session protocol sipv2
  dtmf-relay rtp-nte
  codec g711ulaw byte 160

SIP digest authentication mechanism:

This is required if you have SIP phones on different segments!!

We can’t use ID MAC for phones if they are not connected directly to the CME, because the registration request will come with a different Mac than the phone.

So the other option will be to use the SIP digest authentication mechanism:
Ex using xlite client

 Voice register global
   authenticate register ( this cmd will broke registration of other local phones)

Voice register dn 2
 number 3006

 Voice register pool 2
  id mac 1234.4567.6789
  type 7960
  number 1 dn 2
  dtmf-relay rtp-nte
  description 32-14-3006
  codec g711ulaw
  username phone1 password cisco

Voice register global
  create profile  ( to create cnf file)

Show tftp-binding

More flash:/SIP12345676789.cnf (see username password)
  
Configure username / password on the SIP Phone

Authentication user name: phone1
Password: cisco
user name : 3006
Domain> 10.10.202.1  (CME IP)

SIP IP Phones over WAN:

By default the phone will only trust the packets from IP address configured in:

Voice register global
 source-address 10.10.202.1 
 port 5060

So if messages come from the serial interface of the CME with a different IP the phone will not be able to register.

To resolve this issue we should bind SIPmessages:

Voice service voip
  SIP
  bind all source-interface vlan400 (voice vlan)