|
0:00:12
|
Good morning everyone, welcome to what's effectively day ten.
|
|
0:00:17
|
We're going to be covering high availability and CME
|
|
0:00:24
|
as they really go hand in hand today. It looked like someone was
|
|
0:00:27
|
typing a question and as I go into the full screen presentation
|
|
0:00:32
|
mode for my slides, I won't be able to see any questions.
|
|
0:00:35
|
I will be able to as we've noted in previous classes, online classes
|
|
0:00:40
|
that once I'm out of PowerPoint presentation mode and into
|
|
0:00:44
|
demonstration mode, I'll be able to see anyone's questions
|
|
0:00:47
|
but if you have a question please go ahead or a comment
|
|
0:00:51
|
please go ahead and type that now before we begin.
|
|
0:00:53
|
Otherwise, we'll go ahead and get started.
|
|
0:00:56
|
And once we get started in the lecture, again, just
|
|
0:00:59
|
please hold all questions until -- or you can type anything
|
|
0:01:04
|
but just note that I won't really be able to see them until
|
|
0:01:08
|
we're in demo mode. Ok, so welcome everyone back to
|
|
0:01:11
|
day 10 of INE's CCIE Voice Advanced Technologies class.
|
|
0:01:17
|
Today we're going to be talking about high availability and we'll
|
|
0:01:20
|
also cover unified communication manager express or CME as some
|
|
0:01:25
|
people still call it,
|
|
0:01:27
|
as those two really go hand in hand and so what we'll do
|
|
0:01:30
|
we do have two different slide decks for those. We'll go ahead and
|
|
0:01:33
|
go over high availability. We'll take a break and then we'll go
|
|
0:01:36
|
ahead and talk about the concepts and the slides and
|
|
0:01:39
|
lecture for the CME and then we'll launch into the demonstration for both.
|
|
0:01:43
|
As I mentioned because they really do go hand in hand.
|
|
0:01:47
|
So when we're talking about high availability or SRST
|
|
0:01:51
|
Survivable Remote Site Telephony
|
|
0:01:53
|
as it applies to the lab.
|
|
0:01:55
|
The first thing to note is that if you've been doing deployments
|
|
0:01:58
|
or working with unified communications for any time
|
|
0:02:01
|
especially in the recent past few years, you know that there's also
|
|
0:02:05
|
something called SRSV. Now we're not talking about messaging
|
|
0:02:09
|
today, that should be tomorrow I believe, but
|
|
0:02:12
|
it should just be noted right upfront that while we do have to
|
|
0:02:16
|
deal, contend, learn about and execute Survivable Remote Site Telephony
|
|
0:02:22
|
so the ability to have phones registered or fall back to a
|
|
0:02:27
|
local, typically router based skinny server or SIP server
|
|
0:02:34
|
depending on what kind of phone loads we have.
|
|
0:02:36
|
While there's a WAN down type situation or just if CUCM is not
|
|
0:02:41
|
reachable for whatever reason, we don't have to deal with
|
|
0:02:44
|
Survivable Remote Site Voice Mail in the lab exam, so there's no
|
|
0:02:48
|
unified or universal messaging gateway and while we do have
|
|
0:02:53
|
CUE, Cisco Unity Express, which I mentioned we'll talk about tomorrow
|
|
0:02:57
|
we don't have to deal with it in SRSV mode per se although
|
|
0:03:02
|
there is a way that we can execute that and so we'll talk about that
|
|
0:03:08
|
as an option tomorrow, but as we're dealing with today
|
|
0:03:11
|
Survivable Remote Site Telephony, this really as I began to mention
|
|
0:03:18
|
deals with the ability that if we have a situation where
|
|
0:03:22
|
our Unified Communication Manager, so the big, the CUCM
|
|
0:03:27
|
any of our call processing engines, so whether that means
|
|
0:03:32
|
the WAN is actually down or just the Unified Communication Manager
|
|
0:03:38
|
call processing engines that we happen to be registered.
|
|
0:03:40
|
In our lab, we only have a publisher and a subscriber
|
|
0:03:43
|
so chances are the CCIE lab will have you make both of those
|
|
0:03:47
|
call processing engines. That is to say that they'll have you
|
|
0:03:50
|
enable the call manager service which of course we don't have to
|
|
0:03:56
|
do on all servers in a large cluster, but chances are
|
|
0:03:59
|
they'll have you enable that on both and that they will act as
|
|
0:04:03
|
redundant CPE or Call Processing Engines for one another.
|
|
0:04:07
|
If they should both be down or if the WAN should be down
|
|
0:04:09
|
we need to have some sort of failback or fallback mechanism.
|
|
0:04:16
|
Now the couple different ways that we can do this is we can
|
|
0:04:21
|
have either what's known as traditional SRST, so that is
|
|
0:04:26
|
running on a router, the skinny server that uses the
|
|
0:04:30
|
command call-manager-fallback
|
|
0:04:34
|
and we place all of our commands underneath that and we'll take a look at that today
|
|
0:04:37
|
as one option, traditional SRST as we'll call it.
|
|
0:04:41
|
Or we can also use something which would be known as
|
|
0:04:45
|
CME as SRST, so communication manager express running in
|
|
0:04:51
|
SRST mode and then of course we'll also have the ability to
|
|
0:04:57
|
just execute CME without SRST mode and really
|
|
0:05:00
|
the difference between CME as SRST and CME just plain vanilla
|
|
0:05:07
|
without the SRST mode really isn't that much of a difference.
|
|
0:05:10
|
In fact, it's mainly one command that we'll take a look at that
|
|
0:05:16
|
draws this distinction.
|
|
0:05:19
|
But either way, whether it's running as SRST or without SRST
|
|
0:05:24
|
all of the same functionality is there. It's just that with SRST
|
|
0:05:28
|
we have a little bit of added functionality where the phones
|
|
0:05:33
|
basically as they're falling back to local router control for the
|
|
0:05:40
|
skinny server they, the phones, tell the CME server
|
|
0:05:44
|
basically how they're configured.
|
|
0:05:46
|
And depending on how we implement this SRST command
|
|
0:05:51
|
for CME depends on whether the server basically listens or not.
|
|
0:05:54
|
Listens to all of what they have to say part of what they have
|
|
0:05:57
|
to say or none of what they have to say and if it listens to
|
|
0:06:01
|
only part or none of what they have to say overrides
|
|
0:06:05
|
the appropriate settings.
|
|
0:06:07
|
So that's one of the primary things to note that with fallback
|
|
0:06:11
|
to SRST mode and certainly traditional SRST mode this is
|
|
0:06:15
|
about the only way that -- well, other than a few
|
|
0:06:19
|
things that we can set globally under the call-manager-fallback
|
|
0:06:23
|
subsection, this is the only way that this works
|
|
0:06:26
|
and that is just again to say that the phones instruct
|
|
0:06:30
|
the CME server or the skinny server I should say
|
|
0:06:35
|
as to their configuration. They effectively tell the CME server
|
|
0:06:40
|
what their configuration was when they were registered to
|
|
0:06:42
|
the CUCM in at least as much as they know, so
|
|
0:06:46
|
that is as much information as the individual phones hold
|
|
0:06:51
|
in running memory. They instruct the
|
|
0:06:55
|
CME server of that information.
|
|
0:06:57
|
Now there's certain things that they cannot instruct the server of
|
|
0:07:01
|
such as if they were a part of -- and these are the things
|
|
0:07:04
|
they don't have knowledge of, they being the phones
|
|
0:07:08
|
such as if they were a part of a line group ultimately a hunt group.
|
|
0:07:13
|
These are things and types of functionality that if we needed
|
|
0:07:16
|
to provide, we could probably only do with CME as SRST and not traditional
|
|
0:07:23
|
call manager fallback or traditional SRST.
|
|
0:07:27
|
So taking a look at what we have on the slide, dealing with phone
|
|
0:07:30
|
calling features, phone features, dialing, dial plan, so as I note here
|
|
0:07:36
|
when you were back configuring your phone and call features and
|
|
0:07:39
|
I'm a little bit alluding to something that we'll talk about
|
|
0:07:45
|
on our last day in terms of overall strategy. When we talk
|
|
0:07:50
|
about strategy, one of the things that we'll mention is that we will
|
|
0:07:54
|
encourage you to read the entire exam when you sit down to your
|
|
0:08:00
|
real exam or even when you're practicing a mock lab. Not necessarily
|
|
0:08:04
|
while you're going through ATC or our volume one or technology specific labs
|
|
0:08:09
|
but when you're going through a mock lab, a full eight-hour lab,
|
|
0:08:13
|
or even if you're practicing with a four-hour mock lab, a subset of that
|
|
0:08:18
|
just to gain speed and accuracy, if you're dealing with a self-paced mock lab
|
|
0:08:23
|
or if you're in the real lab, we'll encourage you to read the entire
|
|
0:08:26
|
lab before beginning any sort of configuration
|
|
0:08:29
|
and there's a number of reasons for this, one is to get an overall picture
|
|
0:08:34
|
of what is going to be asked of you, so one of the things -- and really this
|
|
0:08:40
|
has to do with the eye in CCIE, the interworking or internetworking
|
|
0:08:45
|
so really how things relate to one another and what sort of
|
|
0:08:49
|
dependencies become inherent within the requirements of whatever
|
|
0:08:56
|
mock or real lab you're happening to be working on at that moment
|
|
0:09:01
|
that day, so one of those interworkings is dealing with the fallback, traditional or
|
|
0:09:10
|
CME as SRST and then also how it relates to all of the other features
|
|
0:09:18
|
that you were reading about and/or configuring earlier on.
|
|
0:09:23
|
So this is where we're saying when you're back configuring or at least
|
|
0:09:27
|
reading about the phone, features, the calling features and we'll get to
|
|
0:09:32
|
dial plan here in a moment, one of the things you should have been
|
|
0:09:35
|
thinking about and hopefully making a few notes on
|
|
0:09:39
|
nothing terribly verbose, but just a few short hand notes
|
|
0:09:43
|
is about which phones will eventually fall back to SRST
|
|
0:09:47
|
if any, most likely you'll have some phones at some location
|
|
0:09:52
|
and if the exam is actually asking you to match like functionality
|
|
0:09:56
|
or exact functionality for these phones.
|
|
0:10:00
|
Remember the wording on the CCIE exams if fairly specific.
|
|
0:10:05
|
If not, very specific or totally specific.
|
|
0:10:09
|
As always, if you are unclear of the exact wording, ask the proctor
|
|
0:10:15
|
and I know you've said this before, I'll say it again in strategy. Depending on
|
|
0:10:19
|
location or testing facility that you go to, to take and sit your exam
|
|
0:10:24
|
that will have a large part to do with how much the proctors can
|
|
0:10:30
|
help you with the wording, so in other words, if there happens to
|
|
0:10:34
|
be a CCIE Voice proctor there and by that I mean a proctor that
|
|
0:10:38
|
actually has their CCIE Voice and not just some of the other exams
|
|
0:10:43
|
maybe route switch, maybe they have security, maybe they even have
|
|
0:10:47
|
three, route switch security service provider, but they don't have
|
|
0:10:50
|
voice per se, they might not be able to help you with as much
|
|
0:10:54
|
clarification or wording as say a proctor that also has the CCIE Voice.
|
|
0:10:58
|
Now it's not to say they won't be able help you with any, they do have the
|
|
0:11:01
|
grading configuration, so they can take a look at those and probably
|
|
0:11:06
|
deduce or come up with a little bit of helpful information
|
|
0:11:09
|
as to clarification on the wording; however, if you go to a location that
|
|
0:11:15
|
has a CCIE Voice, you're probably going to get a much better response
|
|
0:11:20
|
as that person is probably intimately familiar with the
|
|
0:11:23
|
exam, possibly even helps on the content advisory board as to what
|
|
0:11:27
|
goes into the exam. One of those locations is RTP or
|
|
0:11:32
|
Research Triangle Park Raleigh, North Carolina in the US
|
|
0:11:37
|
and there is a proctor there at least for the time being
|
|
0:11:41
|
she's relatively new at least as of this recording about
|
|
0:11:45
|
a year and a half, two years, so her name is Kelly and she is a
|
|
0:11:49
|
CCIE Voice and she's really helpful, so if you have the chance
|
|
0:11:52
|
that's a good place so that you can get clarification.
|
|
0:11:54
|
But getting back to what you would get clarification of
|
|
0:11:58
|
now that I've done that little side tangent,
|
|
0:12:00
|
this deals with the wording of if you had something in your
|
|
0:12:04
|
exam that said, 'When phones fall back to SRST, provide like functionality
|
|
0:12:12
|
as to when they're registered to CUCM', so during a WAN outage
|
|
0:12:17
|
phones should have like functionality compared to when they're normally
|
|
0:12:23
|
registered to CUCM. If they said something like that
|
|
0:12:27
|
that would cause me as a test taker to compile a list of
|
|
0:12:32
|
all of the different things and we'll go over what those things are
|
|
0:12:36
|
a list of all the different things that I thought I could provide
|
|
0:12:42
|
in terms of like functionality and then I would probably take that up to the
|
|
0:12:47
|
proctor and ask him or her 'Here are all the things that I could come up with
|
|
0:12:53
|
to provide like functionality.' Now if you said something
|
|
0:12:57
|
if you asked a question something akin to 'Did I get everything on the list?'
|
|
0:13:05
|
or 'Am I missing everything?' they're not going to answer
|
|
0:13:08
|
that question, they're not going to do the exam for you
|
|
0:13:11
|
but you might say something like 'Am I on the right track?'
|
|
0:13:16
|
'Are you expecting Mr./Mrs. proctor or Mr./Ms. proctor
|
|
0:13:21
|
are you expecting exact functionality or are you just expecting that I provide
|
|
0:13:29
|
as much as possible?' And again, they're not going to tell you what is included in
|
|
0:13:32
|
as much as possible. That's left up to you and your expert level knowledge
|
|
0:13:36
|
but they hopefully will help you out some there.
|
|
0:13:39
|
If they did happen to use the word exact functionality, well
|
|
0:13:43
|
that's a pretty specific word and obviously it means that anything
|
|
0:13:48
|
that they asked you to do earlier in the exam whether it be
|
|
0:13:53
|
class of restriction, so some of those things that I said I'd mention
|
|
0:13:56
|
class of restriction, dial plan, so the way that digits are dialed
|
|
0:14:01
|
the sounds that are heard, so in other words, if I dialed 9 for
|
|
0:14:06
|
secondary dial tone, I should still hear secondary dial tone
|
|
0:14:09
|
after I dial the first 9, if I dialed zero for secondary
|
|
0:14:12
|
dial tone, I should still hear zero. What I should be able
|
|
0:14:16
|
to dial in terms of how the digits are laid out in terms of
|
|
0:14:21
|
how I might have overlapping dial plans as we talked about
|
|
0:14:26
|
earlier in this ATC module, how I might not, how I might have
|
|
0:14:30
|
terminating characters or interdigit timeout characters
|
|
0:14:34
|
so the hash symbol we configured in CUCM, of course we could have
|
|
0:14:39
|
changed that in CUCM we also can change that -- actually it's
|
|
0:14:45
|
because of the discard digit trailing hash, so predot trailing hash
|
|
0:14:49
|
in CUCM, it's not that we couldn't use another digit, but
|
|
0:14:53
|
hash is probably the digit that we would almost use to terminate
|
|
0:14:58
|
interdigit timeout, not to mention it's the one users are most familiar with
|
|
0:15:02
|
but we do have the ability to change that in CME or in any IOS router.
|
|
0:15:07
|
We can take a look at that, but all of those things
|
|
0:15:11
|
in terms of the features, the mobility, so unified mobility for
|
|
0:15:18
|
instance or single-number reach is something that doesn't make its
|
|
0:15:21
|
way into CME as SRST or just CME in general until 7.1, CME 7.1
|
|
0:15:28
|
which actually doesn't come until -- and if you're unfamiliar
|
|
0:15:31
|
with this, the way that the CME versions are laid out or
|
|
0:15:36
|
dependant, they're directly dependant on the IOS version
|
|
0:15:40
|
that's applied to the router, so for running 124-20 T
|
|
0:15:44
|
which the lab may have you test on, that's the first version
|
|
0:15:48
|
where we had CME 7.0. -- I believe it was 7.0.0
|
|
0:15:57
|
If the lab is running 124-22 T and it could be running
|
|
0:16:02
|
124-20 T or 124-22 T
|
|
0:16:05
|
then that is 7.0.1 CME version 7.0.1
|
|
0:16:10
|
If the lab was running 12.4.24 T
|
|
0:16:16
|
that is where we get CME 7.1
|
|
0:16:20
|
and we know that the lab is testing 7.0
|
|
0:16:23
|
for CUCM, for CME, so we know that we're either going to have
|
|
0:16:27
|
124-20 T or 124-22 T IOS versions, but
|
|
0:16:32
|
not IOS 124-24 T
|
|
0:16:35
|
and so therefore we also know that we're not going to have
|
|
0:16:38
|
CME 7.1, that's where we really started, so we're not going to have
|
|
0:16:42
|
single-number reach or the equivalent in CME is actually
|
|
0:16:48
|
called single-number reach where the equivalent of mobile connect
|
|
0:16:51
|
so that would be one feature that we cannot provide.
|
|
0:16:54
|
It's one of the reasons it's probably unlikely that they
|
|
0:16:56
|
would say exact functionality or if they did, they wouldn't
|
|
0:17:00
|
have made whatever phone was going to fall back, say a Branch 1
|
|
0:17:04
|
or a Branch 2 phone, they wouldn't have given them in the normal CUCM
|
|
0:17:11
|
registration they wouldn't have given them that advanced functionality
|
|
0:17:15
|
like they wouldn't have assigned you or told you to assign that phone
|
|
0:17:21
|
mobile connect or single-number reach functionality.
|
|
0:17:25
|
But things like hunt groups, we cannot provide with traditional
|
|
0:17:30
|
SRST, but we can provide with CME as SRST.
|
|
0:17:34
|
So again, these are perfect things for you to consider the exam
|
|
0:17:39
|
intends or doesn't intend for you to configure, but they might not
|
|
0:17:44
|
specifically call for in the wording of the SRST type task, so this is
|
|
0:17:48
|
where you need to be insightful, look at the wording, if it says
|
|
0:17:52
|
like functionality, they may not say 'Make sure the Branch 2
|
|
0:17:57
|
phone 1 and phone 2 are in a hunt group'; however, if they were
|
|
0:18:02
|
in CUCM, then unless they told you to provide specifically
|
|
0:18:07
|
traditional SRST which is the call-manager-fallback command
|
|
0:18:10
|
which we can't provide a hunt group per se, I guess we could
|
|
0:18:14
|
maybe with BACD, but I would exhaust all options, so maybe I
|
|
0:18:19
|
could with BACD for instance. If it was CME as SRST or if they didn't
|
|
0:18:23
|
tell me which one to use, I would probably pick CME as SRST
|
|
0:18:27
|
and I could provide native hunt group functionality, so also make
|
|
0:18:31
|
sure that dial plan works in the exact same fashion.
|
|
0:18:35
|
I've basically already said this, but I want to touch on this again
|
|
0:18:38
|
and just say that the end goal -- and this is true in real life, real
|
|
0:18:43
|
deployments, although that's not primarily what we're focused on
|
|
0:18:46
|
with this class, it certainly is the goal in real life as well
|
|
0:18:50
|
is that the end user -- the goal is for the end user to
|
|
0:18:55
|
not know anything has happened except that maybe if they happen
|
|
0:18:58
|
to be looking that their phone blipped quickly. Now if they were
|
|
0:19:01
|
actually on the phone in a live conversation unless
|
|
0:19:05
|
that remote site was an MGCP gateway where the PRI is
|
|
0:19:12
|
backhauled -- well the Layer 3 Q.931 D channel is backhauled
|
|
0:19:18
|
and so therefore CUCM is in control of that PRI and so
|
|
0:19:23
|
therefore if the WAN link or communications to CUCM goes
|
|
0:19:27
|
down, then so does the D channel so Layer 3 of our
|
|
0:19:31
|
ISDN voice PRI goes down and then router takes over
|
|
0:19:36
|
with the call application alternate default, so it
|
|
0:19:41
|
takes over with local control of that D channel. Other than that
|
|
0:19:45
|
situation where we would lose all calls, all B channels would drop
|
|
0:19:49
|
their calls because the D channel, the controlled channel, has dropped.
|
|
0:19:55
|
Other than that situation if we had a SIP or H.323 controlled
|
|
0:20:03
|
gateway, even if we were on the phone, we would end up
|
|
0:20:08
|
preserving our call information,
|
|
0:20:12
|
so our RTP stream would continue to be up, we would
|
|
0:20:17
|
continue to be live on the call, we may not have all the
|
|
0:20:22
|
supplementary features like hold, transfer, so that might
|
|
0:20:26
|
become noticeable if I tried to do that, but then
|
|
0:20:28
|
as soon as I hung up and maybe tried to initiate the same or
|
|
0:20:32
|
another call, I would have all those features as the local
|
|
0:20:36
|
router would be in control of that, but again, unless the user is on a
|
|
0:20:40
|
call on an MGCP gateway or on another type gateway
|
|
0:20:44
|
and trying to invoke a supplementary feature or unless they just happen to be
|
|
0:20:48
|
looking at their phones and see a quick blip that is the phone
|
|
0:20:51
|
unregister from the Pub or Sub, try the fallback, can't reach either
|
|
0:20:56
|
and quickly falls back and tries the tertiary call processing server
|
|
0:21:02
|
which is the SRST gateway, then essentially unless any of those happen
|
|
0:21:09
|
it should be a seamless experience and so even if they do fall back,
|
|
0:21:13
|
even if they happen to notice, it should be a seamless experience,
|
|
0:21:15
|
they might think glitch in the system, something happened, but we're back up
|
|
0:21:23
|
and everything looks the same. And this goes down to even the
|
|
0:21:28
|
finite details of what's called the system message at least in
|
|
0:21:32
|
CME and traditional SRST is called the system message at the bottom
|
|
0:21:37
|
of the display, so in CUCM we can't really control what the
|
|
0:21:43
|
bottom of the phone display, so right above the soft keys
|
|
0:21:48
|
what that says by default.
|
|
0:21:51
|
In CME and traditional SRST, we can and it's called the
|
|
0:21:55
|
system primary message or on the older phones the secondary
|
|
0:21:59
|
message, but we don't have to deal with those, so the system
|
|
0:22:02
|
primary message. By default, if left untouched -- it depends on
|
|
0:22:06
|
actually which version of CME but it might say something like
|
|
0:22:10
|
in traditional fallback it might say, 'Call manager fallback operating'
|
|
0:22:16
|
if it's CME, it might say, 'Unified CME' or 'Unified CM Express'
|
|
0:22:24
|
We can of course change it to the same default message and
|
|
0:22:28
|
by default I mean we don't have an MWI, so where it says one or two
|
|
0:22:32
|
or five messages waiting, we don't see one or five missed calls or
|
|
0:22:37
|
whatever the recent situation might have been, the recent status
|
|
0:22:42
|
change might have been to change the default, but the
|
|
0:22:45
|
default message if we were just to register a phone to CUCM for
|
|
0:22:49
|
the first time, the default message we see is 'Your
|
|
0:22:52
|
Current Options' and that's a capital Y and then all of the rest
|
|
0:22:56
|
of the letters including the C in Current and O in Options
|
|
0:23:00
|
'Your Current Options' all the rest of the letters are all lower case.
|
|
0:23:04
|
That would be another thing, a small nuance as would the top
|
|
0:23:09
|
right display which in CUCM we control with the external phone number mask
|
|
0:23:15
|
configured on the first line of the device, that's one more thing
|
|
0:23:19
|
that in CME we would need to configure
|
|
0:23:22
|
and make look the same. Small things that actually
|
|
0:23:26
|
make a big difference psychologically to the user in just having a seamless
|
|
0:23:30
|
experience, everything looks the same.
|
|
0:23:33
|
Now there's many, many features, there's many things in terms of
|
|
0:23:38
|
phone features, calling features whether it be park slots or other
|
|
0:23:43
|
things that are similar to CUCM as well as many other features
|
|
0:23:48
|
that don't exist in CUCM, but do exist in CME or CME as SRST
|
|
0:23:55
|
such as maybe say paging, multicast paging
|
|
0:23:59
|
and if you're fallback because we're talking about high availability here
|
|
0:24:02
|
it's probably unlikely that they're going to have you set up something
|
|
0:24:06
|
like a paging group or night bell, a night service bell
|
|
0:24:10
|
something that doesn't exist in CUCM when the phone is normally
|
|
0:24:14
|
registered, but does exist in fallback mode. It's highly unlikely
|
|
0:24:18
|
they'll have you set something like that up.
|
|
0:24:20
|
But if you we're dealing with CME standalone, then they very well may.
|
|
0:24:25
|
Either way, there are many, many features and we actually go
|
|
0:24:29
|
through them meticulously and cover then exhaustively
|
|
0:24:33
|
in the deep dives, the voice deep dives, and I believe it's under
|
|
0:24:38
|
modules, I think it's -- I believe it's modules 16, 17 and 18
|
|
0:24:46
|
but anyway, we spend almost three full days covering every
|
|
0:24:50
|
single feature. We're not going to cover every single
|
|
0:24:55
|
feature today, what we will do is go over the documentation website
|
|
0:25:00
|
and the administration guide and show you where you can
|
|
0:25:03
|
quickly find all of those. If you want to actually watch
|
|
0:25:06
|
each and every feature individually as I apply them and talk about them and
|
|
0:25:10
|
how they're done, I would say reference the deep dives
|
|
0:25:14
|
but in terms of being prepared for the lab, it might be a good idea to
|
|
0:25:19
|
go over them maybe once, but I wouldn't spend the majority of
|
|
0:25:24
|
my time studying all of those individual features because the
|
|
0:25:29
|
amount of time studying is probably not going to necessarily
|
|
0:25:33
|
yield you a one for one in terms of points on the exam.
|
|
0:25:37
|
Instead, it's very important to at least know that they exist
|
|
0:25:42
|
and know where to find them and like I said, maybe you have
|
|
0:25:45
|
configured them once, but not like some of the other things in
|
|
0:25:48
|
the exam where you'll have really muscle memory and you'll be
|
|
0:25:52
|
able to configure everything in Notepad because you'll be that
|
|
0:25:56
|
readily fluent with the commands for let's say IOS gateways and dial
|
|
0:26:02
|
peers and things like that, voice translation rules
|
|
0:26:04
|
but when it comes to each and every feature in CME, the most
|
|
0:26:08
|
important thing is to like I said maybe you have done it once, but
|
|
0:26:11
|
know where to find them in the universe CD or in the Cisco
|
|
0:26:16
|
documentation website as they call it now, in that documentation
|
|
0:26:20
|
in the one administrator guide which is very easy to get to
|
|
0:26:24
|
and we'll do that collectively here in just a moment and then
|
|
0:26:27
|
be able to grab the entire PDF which you will have
|
|
0:26:31
|
Acrobat Reader in the lab exam on your candidate desktop
|
|
0:26:35
|
and then do either look in the table of contents which is
|
|
0:26:39
|
probably the quickest or you can also do a Control F and find in that
|
|
0:26:46
|
PDF and I recommend looking at the PDF because you can search the
|
|
0:26:50
|
entire document, if you need to do a Control Find, Control F
|
|
0:26:53
|
rather than the web page where each individual subsection, each
|
|
0:27:00
|
or I should say each individual feature is a sub section, it's another
|
|
0:27:05
|
page that you have to click on and then do a Control Find, so
|
|
0:27:08
|
you might find yourself having to click Control Find 25 times for
|
|
0:27:12
|
25 different pages as opposed to the PDF where you can do it once.
|
|
0:27:15
|
So taking a look back in CUCM
|
|
0:27:20
|
it might be the case that in dealing with phones falling back to SRST
|
|
0:27:24
|
let's say at Branch 1 and/or Branch 2 and we'll take a look at both.
|
|
0:27:30
|
It might be the case that the CUCM and typically is
|
|
0:27:33
|
the case that the CUCM Publisher and Subscriber or our CPE devices
|
|
0:27:39
|
are still in an operating mode for those that still have connectivity,
|
|
0:27:45
|
WAN connectivity or LAN connectivity.
|
|
0:27:47
|
So certainly for the corporate headquarter phones or the main
|
|
0:27:53
|
site phones and so these phones need to be able to
|
|
0:27:58
|
still do four-digit dial that is their experience
|
|
0:28:02
|
needs to be seamless, they shouldn't at least the end user
|
|
0:28:05
|
shouldn't know that anything's happened to the Branch 1 and/or
|
|
0:28:09
|
Branch 2 site phones and if the phones at corporate headquarters
|
|
0:28:13
|
go to dial the four-digit extension of a Branch 1 or Branch 2 phone,
|
|
0:28:17
|
they should be able to reach it seamlessly. Now the same would be
|
|
0:28:21
|
true and should be said, in fact was implicitly said
|
|
0:28:25
|
on the last slide, in terms of Branch 1 or Branch 2 phones
|
|
0:28:28
|
being down and registered in their fallback SRST mode
|
|
0:28:32
|
they also should have four-digit dial and we will cover that as part
|
|
0:28:35
|
of the dial plan in the demonstration when those phones fall back.
|
|
0:28:39
|
But back to this, the corporate headquarter phones that are
|
|
0:28:42
|
registered to the CUCM, they should have four-digit dial and
|
|
0:28:47
|
actually that's the next slide sorry, that's where we have
|
|
0:28:50
|
call forward on unregister. Now when we talked about
|
|
0:28:55
|
globalization earlier and we talked about when and if we're using
|
|
0:29:00
|
a fully globalized dial plan, how we get call forward on unregister
|
|
0:29:08
|
for free, so the idea is that all call forward destinations whether they
|
|
0:29:13
|
be call forward all, call forward busy, call forward -- and call forward
|
|
0:29:17
|
all might not necessarily be because the user might enter it from the keypad
|
|
0:29:21
|
so that might be the exception, but call forward busy, call forward
|
|
0:29:24
|
no answer and specifically what we're talking about here
|
|
0:29:27
|
call forward on unregister also known as CFUR, those should
|
|
0:29:32
|
be in the standardized globalized plus E.164 format.
|
|
0:29:38
|
Of course the calling search space for that call forward on
|
|
0:29:42
|
unregister determines at least its reachability to the PSTN pattern
|
|
0:29:48
|
the plus PSTN pattern which we have configured as translation patterns
|
|
0:29:52
|
and determines the routing path unless we're dealing with
|
|
0:29:57
|
an SLRG so a Standard Local Route Group.
|
|
0:30:00
|
Ok, in that case calling party determines the path unless we
|
|
0:30:04
|
have what we had already configured back in the dial plan
|
|
0:30:09
|
section which was tail end hop-off and then the tail end hop-off
|
|
0:30:12
|
should point those to the proper egress gateway.
|
|
0:30:17
|
And then of course, the egress called party transformation patterns
|
|
0:30:20
|
take care of all the rest.
|
|
0:30:22
|
So for the Branch 1 or Branch 2 phones, our CFUR will have essentially
|
|
0:30:27
|
their plus and then full E.164 number
|
|
0:30:30
|
with their specific DID, so we'll typically use masks for these so that we can
|
|
0:30:34
|
deploy them quickly and across all phones in that Branch site
|
|
0:30:39
|
using the BAT tool, the Bulk Administration Tool
|
|
0:30:41
|
and then as we've demonstrated before, egress called party transformation
|
|
0:30:46
|
patterns take care of all the rest and that's where we
|
|
0:30:50
|
say call forward on unregister for free.
|
|
0:30:52
|
When we say for free, we basically mean that in
|
|
0:30:57
|
in previous implementations of CUCM, we had to be very specific
|
|
0:31:02
|
as to how that number was formatted and it was very much
|
|
0:31:09
|
dependant on which gateway it would go out, so this makes it
|
|
0:31:13
|
sort of a -- I don't know about brainless task, but
|
|
0:31:16
|
a very low CPU cycle task in terms of your brain
|
|
0:31:20
|
and what you have to think about.
|
|
0:31:23
|
Going back to the last slide which I had
|
|
0:31:26
|
dealing with another high availability situation, so
|
|
0:31:30
|
not a WAN down situation where we have to deal with
|
|
0:31:33
|
call forward on unregister for the phones that are still
|
|
0:31:39
|
up and registered to our CPE to be able to reach the phones that are
|
|
0:31:43
|
down or in fallback and then also SRST for the phones that are in fallback
|
|
0:31:48
|
to be able to be registered and dial.
|
|
0:31:50
|
But when we have a not enough bandwidth situation
|
|
0:31:55
|
so we have some form of call admission control
|
|
0:31:58
|
whether that's standard, traditional, locations based call admission control
|
|
0:32:04
|
or whether that's RSVP enabled locations, it really doesn't matter
|
|
0:32:09
|
so it doesn't matter whether CUCM is determining that there's not
|
|
0:32:12
|
enough bandwidth or CUCM is invoking an RSVP agent on an IOS
|
|
0:32:16
|
router and the IOS router running RSVP signals back to CUCM
|
|
0:32:22
|
via skinny and the RSVP agent that there's not enough bandwidth
|
|
0:32:26
|
but in some fashion call admission control says there's not enough
|
|
0:32:30
|
bandwidth to continue and the first thing that's asked is
|
|
0:32:34
|
Is Automated Alternate Routing or AAR, is it enabled?
|
|
0:32:41
|
So it needs to be enabled in the CUCM service parameters
|
|
0:32:47
|
for specifically for the CCM service and it's almost at the very bottom
|
|
0:32:52
|
we'll go there and do that. It first needs to be enabled
|
|
0:32:56
|
and then we need to have AAR groups and AAR CSSs
|
|
0:33:02
|
set up for all the devices.
|
|
0:33:05
|
Well, in terms of making it free, globalization or localization
|
|
0:33:10
|
makes this free, so what do we mean by this?
|
|
0:33:14
|
Again, as long as we're using the plus E.164 globalized format
|
|
0:33:21
|
of any of the phones that are to be reached, so their full DID
|
|
0:33:25
|
and this is assuming in a real world environment that they would
|
|
0:33:28
|
actually have a DID that either matches their exact number or
|
|
0:33:32
|
we have some sort of mask or translation for that.
|
|
0:33:35
|
In a lab environment, it's very easy to make sure that
|
|
0:33:37
|
our carrier essentially gives us enough DIDs and that we match them up with the
|
|
0:33:43
|
end point extensions, but as long as we put that full globalized DID
|
|
0:33:49
|
or E.164 number in the either the external phone number mask
|
|
0:33:54
|
field is what's used traditionally, but if that is not desired to be used
|
|
0:34:02
|
in other words, what we want to show in the top right of our
|
|
0:34:05
|
phone displays is not a plus number or is not what we want to
|
|
0:34:10
|
use for AAR, there is a new field on each line, so again, a line is a
|
|
0:34:16
|
DN that's applied to a device, so as we click on that line
|
|
0:34:19
|
there is something called the AAR destination mask
|
|
0:34:25
|
and if we put the plus number in there and then on the device
|
|
0:34:30
|
itself, we give it an AAR CSS as well as an AAR group
|
|
0:34:37
|
and we should put the AAR group on the line as well.
|
|
0:34:40
|
Then what happens is the calling party is trying to reach
|
|
0:34:46
|
the called party obviously, the called party is not available
|
|
0:34:50
|
because call admission control reports that there's not enough
|
|
0:34:54
|
bandwidth so CUCM grabs the called party's external
|
|
0:35:01
|
phone number mask or if configured, the AAR destination
|
|
0:35:04
|
mask which is what we should use and then it grabs the
|
|
0:35:10
|
both the calling party and called party's AAR group
|
|
0:35:14
|
so there's a matrix and we basically say for calling and
|
|
0:35:19
|
called should be this AAR group, this dial prefix
|
|
0:35:23
|
and we used to have to set up all of these different dial prefixes
|
|
0:35:27
|
depending on where you were coming from and possibly where
|
|
0:35:29
|
you were going to, but with globalization, this is where we
|
|
0:35:34
|
talk about it making it -- it makes it free.
|
|
0:35:37
|
And the idea is that we need one AAR group cluster wide
|
|
0:35:41
|
and the AAR group no longer needs any prefix and that's the
|
|
0:35:44
|
reason why we can have one group cluster wide, so
|
|
0:35:48
|
for every single device, whether it's a gateway, whether it's a
|
|
0:35:54
|
CTI route point, whether it's a CTI port, whether it's a
|
|
0:36:01
|
hunt pilot, whether it's a Voice mail port,
|
|
0:36:05
|
whether it's a phone device or line, it doesn't matter what it is
|
|
0:36:09
|
they all use the exact same AAR group which has no
|
|
0:36:13
|
prefix and the whole premise of being able to do that is that
|
|
0:36:17
|
we have the plus format of the numbers that is they
|
|
0:36:19
|
don't need changed depending on where the call is coming from
|
|
0:36:26
|
and where it's going to
|
|
0:36:29
|
because as we've said, we've got our plus PSTN translation
|
|
0:36:33
|
patterns, they match those and then the egress gateway's
|
|
0:36:36
|
called party transformation patterns handle all of the
|
|
0:36:40
|
rest of the localization, so again, just as a refresher
|
|
0:36:44
|
what do we mean by that localization? Well, on outbound
|
|
0:36:47
|
to the PSTN calls, those egress gateways handle changing the plus
|
|
0:36:53
|
globalized format of the called number into whatever
|
|
0:36:57
|
that local carrier is expecting.
|
|
0:37:01
|
If we see that the plus -- what follows the plus is
|
|
0:37:05
|
our egress, so we're imagining we are a particular egress gateway
|
|
0:37:09
|
let's say we're corporate headquarter gateway
|
|
0:37:13
|
if we see that the plus is followed by a 1, then we know that
|
|
0:37:17
|
that country code is the US/Canada country code and we know that, that
|
|
0:37:22
|
gateway or the call, the called number, but that that call in progress is
|
|
0:37:27
|
not an international call.
|
|
0:37:29
|
If we see that it's +1206, then we know that it's also not
|
|
0:37:34
|
a national call. It's a local call. Why? Because we've already set up
|
|
0:37:39
|
those called party transformation patterns.
|
|
0:37:43
|
If we saw that it was +1 anything other than 206
|
|
0:37:46
|
we know it's national and if we see it's plus and followed by any
|
|
0:37:50
|
other digit other than a 1, we know it's international
|
|
0:37:52
|
and we know how to format it properly or localize it so that
|
|
0:37:58
|
our local carrier responds properly to that local
|
|
0:38:03
|
national or international call.
|
|
0:38:06
|
Also one other thing to mention real briefly with AAR, if the actual
|
|
0:38:10
|
gateway that's going to be processing this AAR call
|
|
0:38:14
|
typically at the local site, but if you're ever in question, you can
|
|
0:38:18
|
uncheck this or untick this box and see if it makes the
|
|
0:38:23
|
previously failed AAR call begin to work.
|
|
0:38:28
|
Specifically there's a bug with the current 124-20 and 124-22 T
|
|
0:38:35
|
IOS SIP implementation in terms of -- actually, I'm sorry
|
|
0:38:42
|
it's not with those, it's actually a bug -- I apologize, it's actually a bug
|
|
0:38:46
|
with CUCM 7.0.1, I believe this was fixed in 7.1, but the lab
|
|
0:38:52
|
of course is testing CUCM 7.0.1
|
|
0:38:55
|
and it's the fact that CUCM sends a SIP message with an
|
|
0:38:59
|
invalid redirect field and so if we're sending this AAR
|
|
0:39:07
|
call either out or potentially in because we typically are going
|
|
0:39:13
|
out one PSTN gateway and in another to effectively reroute the
|
|
0:39:20
|
call through the PSTN and bypass the WAN.
|
|
0:39:22
|
If we're going out or in a SIP gateway and again in CUCM
|
|
0:39:27
|
there's only a SIP trunk, but when I speak of a gateway I'm typically referring to
|
|
0:39:31
|
us using a SIP trunk in CUCM to control a PRI TDM circuit
|
|
0:39:39
|
I refer to that as a gateway, then we need to deal with that
|
|
0:39:43
|
invalid SIP redirect because otherwise, it would cause the
|
|
0:39:47
|
call to fail and if there's enough bandwidth for it to
|
|
0:39:51
|
fall back to another site, it might work, but if in the case
|
|
0:39:56
|
let's say our corporate headquarter gateway is SIP and
|
|
0:39:58
|
we're redirecting from Branch 1 phone out through the Branch 1
|
|
0:40:02
|
gateway and in the corporate headquarter SIP gateway
|
|
0:40:05
|
then it's going to fail regardless so what we need to do is
|
|
0:40:08
|
on the -- first of all we could untick in the gateway
|
|
0:40:13
|
we could actually untick the remote party ID for that SIP
|
|
0:40:17
|
trunk in CUCM, but then of course we're going to lose all calling
|
|
0:40:22
|
name and caller ID functionality and the other thing, the better
|
|
0:40:28
|
solution to work around this bug is to simply untick
|
|
0:40:31
|
the retain and you see it as the bottom in double quotes here
|
|
0:40:37
|
in the bottom or sixth bullet point untick the retain this
|
|
0:40:44
|
destination in the call forwarding history under the Phone> Line
|
|
0:40:48
|
and AAR Settings
|
|
0:40:52
|
Also one other thing commonly at least in the 7.0 and older
|
|
0:40:58
|
documentation, Cisco refers to AAR not working with CTI route
|
|
0:41:04
|
points and CTI ports, really CTI applications.
|
|
0:41:08
|
And this is sort of true and it's sort of not.
|
|
0:41:11
|
In the later 8.0, 8.5, 9.0 documentation they actually
|
|
0:41:17
|
refer to this a lot more accurately and they've sort of cleared up the
|
|
0:41:22
|
misconception and effectively it is that if a CTI route point
|
|
0:41:27
|
or a CTI port does not show as registered to CUCM, that is
|
|
0:41:31
|
someone created a CTI port or a CTI route point and this was
|
|
0:41:36
|
actually done a lot more in the past before we had the ability
|
|
0:41:39
|
to have DNs that were not applied to a device, so therefore
|
|
0:41:43
|
just a standalone DN or use as a phantom DN
|
|
0:41:48
|
in -- before we had that ability, we used to create
|
|
0:41:51
|
CTI route points to just use as DNs maybe just for a
|
|
0:41:55
|
sort of a dummy depending on your old PBX language
|
|
0:41:59
|
they were either called dummy DNs or phantom DNs
|
|
0:42:02
|
that is they were DNs that weren't applied to anything
|
|
0:42:05
|
they were just used as routing points or forwarding points.
|
|
0:42:08
|
So if you still see those or any other -- a CTI route point
|
|
0:42:14
|
and use as one of those or you see any other reason that you have a CTI route point
|
|
0:42:18
|
or port that is not registered, then those will not work for
|
|
0:42:23
|
Automated Alternate Routing, AAR
|
|
0:42:25
|
but if the application is actually registered, so that is
|
|
0:42:29
|
there actually is some sort of third party computer or even
|
|
0:42:34
|
Cisco computer, but computer telephony integration some sort of
|
|
0:42:38
|
CTI application on the far side that has utilized, has a user name
|
|
0:42:43
|
that in our -- of course our application users in CUCM
|
|
0:42:48
|
that user name is device associated to both the CTI route point and
|
|
0:42:54
|
all the CTI ports and so therefore it has registered those ports to
|
|
0:43:00
|
CUCM or it's registered to CUCM causing those ports to show up
|
|
0:43:05
|
as registered, then AAR will work for those, so some examples
|
|
0:43:08
|
are let's say UCCX or Unified Contact Center Express.
|
|
0:43:13
|
If we have Unified Contact Center Express, it uses
|
|
0:43:17
|
of course CTI route points and ports and it registers those to the actual
|
|
0:43:22
|
CCM server. They show up as registered, AAR will work.
|
|
0:43:27
|
Another thing we already talked about was IPMA.
|
|
0:43:30
|
IPMA is a Tom Cat service running on the call manager
|
|
0:43:33
|
but it registers those CTI route point and CTI ports.
|
|
0:43:38
|
So AAR will work.
|
|
0:43:41
|
Again, it's if everything is configured properly.
|
|
0:43:44
|
In other words, the AAR group is placed every single place
|
|
0:43:48
|
we see in AAR group, we put that one global one there.
|
|
0:43:53
|
AAR calling search space is placed everywhere and the
|
|
0:43:58
|
fully globalized E.164 number is placed
|
|
0:44:00
|
everywhere that we see either an external phone number mask
|
|
0:44:04
|
or an AAR destination mask.
|
|
0:44:06
|
One more thing to keep in mind with AAR is its interactions with
|
|
0:44:10
|
TEHO or Tail End Hop-Off
|
|
0:44:12
|
if say I needed to place a call from corporate headquarters
|
|
0:44:17
|
to Branch 2 and when I try to dial the phone, I would get
|
|
0:44:23
|
not enough bandwidth as a system message and by the way,
|
|
0:44:28
|
there's two system messages that you might see at the bottom of
|
|
0:44:31
|
your phone display screen, of your calling phone display screen.
|
|
0:44:35
|
Seeing only one of those helps with trouble shooting
|
|
0:44:38
|
or seeing both of them helps with troubleshooting.
|
|
0:44:40
|
If I only ever see the message 'not enough bandwidth'
|
|
0:44:45
|
or if I've changed that message in service parameters whatever
|
|
0:44:48
|
I see it changed to, then that means AAR is not invoked
|
|
0:44:53
|
it's not trying to work
|
|
0:44:55
|
or something's misconfigured and it's not even trying to work.
|
|
0:44:59
|
So if I hear -- if I see 'not enough bandwidth'
|
|
0:45:02
|
then I hear reorder tone, then AAR is either not configured
|
|
0:45:06
|
so it's not enabled in service parameters or it's just not set up properly.
|
|
0:45:11
|
If I see 'not enough bandwidth' possibly briefly
|
|
0:45:15
|
sometimes you won't see it at all. If I either see it
|
|
0:45:19
|
first and then I see the message which is 'network
|
|
0:45:23
|
congestion, rerouting' or whether I never see the first message
|
|
0:45:28
|
'not enough bandwidth' sometimes you'll briefly see it, sometimes you won't
|
|
0:45:31
|
just depends on the phone and CUCM's busyness
|
|
0:45:34
|
but if I only see 'network congestion, rerouting'
|
|
0:45:37
|
either one, if I finally see that message, I know that
|
|
0:45:42
|
AAR is enabled and it's attempting to work, but if I also still hear reorder
|
|
0:45:45
|
then I know that I just have something misconfigured
|
|
0:45:50
|
probably the fact that I maybe don't have the ability to match
|
|
0:45:56
|
a translation pattern or a PSTN pattern based on my AAR
|
|
0:46:00
|
calling search space or possibly what we're getting ready to
|
|
0:46:05
|
talk about which is the interaction with tail end hop-off.
|
|
0:46:08
|
So if I'm calling from corporate headquarters over to Branch 2
|
|
0:46:12
|
and I can't reach Branch 2 because there's not enough
|
|
0:46:14
|
bandwidth so I see 'network congestion, rerouting'
|
|
0:46:18
|
If I had tail end hop-off set up and we saw how we did that with the
|
|
0:46:22
|
globalized route pattern that was just a more specific
|
|
0:46:26
|
so we had \+! for our single globalized route pattern that
|
|
0:46:32
|
essentially pointed to the route list of standard local
|
|
0:46:35
|
route group. If I also have a +31, so \+31 bang or exclamation
|
|
0:46:45
|
and that points to just the Branch 2 route list/route group
|
|
0:46:54
|
whether it has backups or not, the idea is that primarily
|
|
0:46:59
|
when I dialed Branch 2 let's say I dialed 3002
|
|
0:47:03
|
as the DN, it wasn't available, so I grabbed the AAR destination
|
|
0:47:08
|
mask of the 3002 phone which was hopefully in the
|
|
0:47:14
|
globalized format it was +31207033002
|
|
0:47:20
|
and then I grabbed the AAR group prefix which is blank
|
|
0:47:23
|
and I tried to match it, match as a translation or PSTN pattern
|
|
0:47:28
|
that forwards on, you know, applies any calling party transformation
|
|
0:47:32
|
make sure class of restrictions works and actually let me just
|
|
0:47:36
|
take one quick moment to say class of restriction should
|
|
0:47:41
|
be looked at as the permissions that the original calling party
|
|
0:47:46
|
had when they were dialing just the 4-digit internal extension
|
|
0:47:50
|
so if I'm dialing a 4-digit extension, as long as I in normal
|
|
0:47:55
|
operations have the ability to dial four digits, I have the
|
|
0:48:00
|
ability to dial that particular internal DN that I was attempting to dial
|
|
0:48:04
|
I should also be able to dial to their AAR, so that is my
|
|
0:48:10
|
AAR CSS should have the ability to dial their DN
|
|
0:48:14
|
regardless of whether it's a subscriber, national or international
|
|
0:48:19
|
call even if my traditional calling doesn't have that right.
|
|
0:48:23
|
So in other words, if I only had the class of restriction to dial
|
|
0:48:29
|
local numbers and it's an international number.
|
|
0:48:34
|
I'm dialing 3002, I'm dialing an internal number as far as I know
|
|
0:48:39
|
the calling party. The system has to redirect that as an international
|
|
0:48:44
|
call per AAR. I should have the AAR calling search space
|
|
0:48:49
|
that would allow me to match and dial that or effectively match
|
|
0:48:53
|
and route that call. I'm not dialing internationally, I'm dialing
|
|
0:48:58
|
internally. The system just happens to be rerouting internationally.
|
|
0:49:01
|
Now that might not be in compliance with a policy that
|
|
0:49:05
|
you or a client of yours might want to set up and that's
|
|
0:49:08
|
perfectly fine, you negotiate that with your in-house staff
|
|
0:49:12
|
policy makers, your clients, whomever
|
|
0:49:14
|
but in terms of the lab, if you were dialing a 4-digit internal extension
|
|
0:49:20
|
and it needs to reroute nationally or internally, that call should work
|
|
0:49:25
|
because from the user's -- the calling user's perspective
|
|
0:49:28
|
they dialed an internal call.
|
|
0:49:30
|
Anyway, getting back to it, so it reroutes or attempts to reroute
|
|
0:49:35
|
that call or tail end hop-off and forward that call out
|
|
0:49:39
|
the Branch 2 gateway, the gateway local to where that
|
|
0:49:44
|
dialed number was. Well the problem was if I didn't have
|
|
0:49:46
|
the ability to dial the Branch 2 phone in the first place because
|
|
0:49:52
|
there wasn't enough bandwidth, then I'm probably not going to
|
|
0:49:56
|
have the or I shouldn't have the ability to dial out that
|
|
0:50:01
|
far end gateway because it should be in the same location.
|
|
0:50:05
|
Ok, again, regardless whether we're using traditional locations
|
|
0:50:10
|
where CUCM controls the bandwidth or RSVP based
|
|
0:50:13
|
locations where the routers control the bandwidth, we
|
|
0:50:16
|
still set up locations in CUCM, we either just set them up
|
|
0:50:20
|
with bandwidth or unlimited bandwidth and list RSVP mandatory
|
|
0:50:25
|
video desired respectively.
|
|
0:50:29
|
Each device still has locations set, so my phones at Branch 2
|
|
0:50:35
|
and my gateways at Branch 2 should all be a part of the same location
|
|
0:50:39
|
as should my phones and gateways at headquarters
|
|
0:50:42
|
and respectively all the other devices at headquarters
|
|
0:50:45
|
and anything at any site, so whether that's a transcoder
|
|
0:50:48
|
or an MOH server etc.
|
|
0:50:53
|
But if I didn't have enough to dial the phone, I'm not going to
|
|
0:50:54
|
have enough to dial out that far end gateway, so
|
|
0:50:57
|
if I only had the far end gateway as an option
|
|
0:51:01
|
then tail end hop-off would fail. Now why would be the or what
|
|
0:51:04
|
would be the reason that I only have the far end gateway
|
|
0:51:07
|
as an option? Well back in my dial plan section if I was instructed
|
|
0:51:11
|
and again, this gets into the interworkings, the dependencies,
|
|
0:51:17
|
the difficulty of the lab -- if back in my dial plan section
|
|
0:51:22
|
I was told to create a tail end hop-off strategy or scenario
|
|
0:51:26
|
whereby corporate headquarter phone's dialing out to a Branch 2
|
|
0:51:33
|
phone number would dial out that remote Branch 2 gateway
|
|
0:51:38
|
and I was told -- I was given the keywords or buzzwords or
|
|
0:51:42
|
something that effectively said only to dial out that gateway
|
|
0:51:47
|
if they were dialing a Branch 2 number.
|
|
0:51:49
|
What they don't mean is in AAR mode, they only mean in traditional
|
|
0:51:53
|
mode where there's enough bandwidth because it's not
|
|
0:51:57
|
possible in AAR mode, so if they had limited me
|
|
0:52:00
|
to only being able to dial out that gateway, then I would
|
|
0:52:04
|
need to make sure that my AAR calling search space
|
|
0:52:07
|
of my corporate headquarter phones or devices could not
|
|
0:52:13
|
see -- and I would do this through calling search spaces
|
|
0:52:16
|
and partitions, but could not see the tail end hop-off pattern
|
|
0:52:21
|
for those remote Branch 2 phone numbers
|
|
0:52:26
|
or prefixes.
|
|
0:52:29
|
However, if back in my dial plan section I was told to do
|
|
0:52:32
|
some sort of tail end hop-off, but I was not given the restriction
|
|
0:52:36
|
that is, they didn't mention it at all that I could only
|
|
0:52:40
|
dial those tail end hop-off numbers out the Branch 2 site.
|
|
0:52:43
|
If they didn't mention it at all, then I would just simply make sure
|
|
0:52:47
|
that my Branch 2 route group because my tail end hop-off
|
|
0:52:54
|
route pattern is going to point to a route list is going to have a
|
|
0:52:56
|
route group that either my Branch 2 route group had
|
|
0:52:59
|
backup gateways in it with the distribution algorithm
|
|
0:53:02
|
of top-down or else that my route list that was used by
|
|
0:53:07
|
that tail end hop-off route pattern, that route list would have
|
|
0:53:10
|
a backup route group of my local, so if you weren't told to only
|
|
0:53:14
|
route out the far, then the simplest answer is just to add a backup
|
|
0:53:18
|
gateway, backup route group. If you were told, then you need to
|
|
0:53:21
|
get a little more creative with your CSSs and partitions to make
|
|
0:53:24
|
sure that it doesn't -- your AAR calling search space from your
|
|
0:53:28
|
calling party does not see that tail end hop-off pattern.
|
|
0:53:32
|
Otherwise, you'll have the inherent impossibility
|
|
0:53:36
|
and you will have reorder tone.
|
