Overview:
The goal of this document is to identify
all the different service interruptions associated with high-speed data (HSD). Each
service interruption will be explained in sufficient detail that it can be understood.
Additionally, each service interruption will be analyzed in terms of what causes the
interruption, what indicators we currently have to detect these interruptions, and what
results from these interruptions we typically experience. Through understanding the
different service interruptions, it is believed that a broadband operator can better
identify the severity of each interruption and what needs to be done to properly address
it. This information will steer the development of monitoring requirements.
Background:
Broadband Service Providers (BSP) have elevated the importance of
monitoring as a component of providing quality service. With multiple services all sharing
the same medium, BSPs must ensure that all facets of the delivery system are kept in
check. This requires that BSPs understand:
- What each service requires
to run optimally
- How each service interacts
with one another
- What triggers loss in
performance and/or service interruption
Monitoring is a means of gathering information about something.
The frequency and scope of information gathered depends greatly on what is being monitored
and the reason it is being monitored. The reasons for monitoring something might include:
- Watching for change
(improvement or deterioration )in something, perhaps beyond certain thresholds
- Watching for the absence or
presence of something
- Watching for some specific
event or sequence of events to take place
As a result of monitoring, one might hope to achieve on of the
following:
Service Interruptions:
The types of outages ranked in order of small to large in terms
of the scale at which it would effect customers are as follows:
- Customer premise
- Node segment
- Node-wide
- Network
- Server
- Third party
Customer Premise Outage
A customer premise outage (Figure 1.0) only impacts a single
customer. Causes of this type of outage include:
- Software installation
conflicts
- PC re-configuration
- Hardware failure (Cable
Modem, Network Interface Card, or CPE)
- Cable (Coax, 10BaseT, or
USB) problems (cut or damaged connector)
- Errors rewiring or splicing
cables
- Downed, damaged, or cut
cable drops
- Power failure
Figure 1.0 Customer Premise
Outage
The following indicates this type of outage:
- Customer call reporting
outage
- Cable modem not pingable
- CPE not pingable
- No Television reception
The results of this type of outage can range from one of the
following (from bad to worse):
- Loss of customer confidence
in the service – based on how quickly they are brought back on line
- Cost of service call to fix
loss of service caused by customer
- Potential impact on other
customers due to ingress caused by drop or customer’s improper wiring
Node Segment Outage
A node segment outage (Figure 1.1) impacts the portion of the customers
connected beyond the location of the outage. Causes of this type of outage
include:
- Hardware failure (mainly
active components such as amplifiers) but could also be passive components like taps,
direct couplers, splitters, etc.
- Signal loss (damaged or cut
plant)
- Localized power failure
Figure 1.1 Node Segment
Outage
The following indicates this type of outage:
- Customer service calls
(takes some minimum number before it is considered a potential outage)
- ChetaNet alert – if
outage was caused by the loss of some active component
- Pathtrack –
considerable swing in network performance (good or bad) – may require large portion
of the plant to be out before pathtrack can sense the outage in this fashion
The results of this type of outage can range from one of the
following (from bad to worse):
- Outage causes ingress which
effects other customers (areas of the plant) and makes the problem more difficult to
locate
- Need to involve a third
party (power company) to fix problem
- Loss of customer confidence
in the service – especially since it is out of their control
- Credits may be required in
some cases if the problem takes considerable time to locate and fix – note that if
this happens at night it may not be fixed till the next morning due to risks associated
with repairs on elevated cables at night.
- Loss of customers seeking
more reliable service – business customers especially
Node-Wide Outage
A node segment outage (Figure 1.2) impacts the portion of the
customers connected to a specific fiber node. Causes of this type of outage include:
- Hardware failure (i.e.
transmission equipment, Cable Modem Termination System, etc.)
- Signal loss (damaged or cut
plant)
- Localized power failure
Figure 1.2 Node-Wide Outage
The following indicates this type of outage:
- ChetaNet would alarm that
node is off line
- Pathtrack would alert
something similar
- Customers would call in
with outages
The results of this type of outage can range from one of the
following (from bad to worse):
- Difficulty finding the
outage – since locating requires a cable cut to be visible the time it takes to
determine this could be lengthy since plant can span miles some of which may not be 100%
visible (night time outages can further complicate the locating exercise)
- Loss of customers
- Bad press
Network Outage
A network outage (Figure 1.3) impacts the customers connected to the
troubled network segment. Causes of this type of outage include:
- Hardware failure (CMTS,
Router, etc.)
- Loss of link (cut or
damaged cable, bad transmission cards)
Figure 1.3 Network Outage
The following indicates this type of outage:
- Network management alert on
link
- Increased or no traffic on
redundant links
- Customer complaints
The results of this type of outage can range from one of the
following (from bad to worse):
- Slower Internet service
- Higher dependency on
alternate paths (in a number of cases there are only two routes to a network) creating a
single point of failure
Server Outage
A server outage (Figure 1.4) impacts all customers who require
the use of the service hosted by the server. Causes of this type of outage include:
- Maintenance problem (file
system overload, operating system problem, etc.)
- Application failure
- Hardware failure
Figure 1.4 Server Outage
The following indicates this type of outage:
- Network operations center
(NOC) alert on hardware failure
- Customer complaint
The results of this type of outage can range from one of the
following (from bad to worse):
- Restart of application
- Replace troubled hardware
(disk drive, NIC, etc) and configure
- Hardware swap – timely as there is no redundant hardware in place
and before new hardware can work it must be properly configured to
support the application
Third Party Outage
A third party outage can impact various numbers of customers
(from 1 to all). Causes of this type of outage include:
- Faulty wiring (poor
workmanship of contractor)
- Failed communications
between broadband operator and contractor
- IP address pool exhausted
- Unplanned outages or down
times that were not coordinated
The following indicates this type of outage:
- Network operations center
(NOC) alert on hardware failure
- Failed installs in area
where other customers are working
- Customer complaint
The results of this type of outage can range from one of the
following (from bad to worse):
- Rescheduled installs or
customer cancel install
- Delays in deployment of new
service due to replacement of contractor
Can Birds-Eye.Net help you or your Company?
Receive your Birds-Eye.Net articles and white
papers hot off
the presses by adding our RSS feed to your reader.