...
The information presented by this report can be used for analysis at multiple levels, such as:
- Troubleshoot or run root cause analysis at a glance.
- Validate IP Network Region design.
- Ensure Gate Keeper resources are working within guidelines.
- Detect and raise alarms at an application level, rather than infrastructure level.
- Use as an indicator of potential network problems.
Identifying traffic trends and the associated impact on IP Signaling capacity can mitigate otherwise unseen issues.
...
The example below shows a distribution of H.323 registration across all Gate Keepers. From this it is possible to validate crucial aspects of the design:
- Confirms that the Gate Keepers are loaded within Avaya recommended tolerances.
- Confirms that the local gatekeeper preference allocation is working as designed.
- A quick count for each gatekeeper confirms that there are no issues with access lists or firewalls.
In the well-engineered example abovebelow, the H.323 traffic on these Gate Keepers is consistently within the Avaya recommended loads.
By setting a high water mark, Availability Manager can be used to try and resolve issues proactively when registrations step above the recommended maximum.
If Availability Manager cannot resolve the issue, then IT support staff are notified by SMS or email, significantly reducing response times, and very possibly likely avoiding a potential outage.
Example 2
In the example below, there has been an outage over the period of the report. A large number of H.323 stations on all gatekeepers deregistered for several hours between 04:00 AM and 11PM on the 12th and 12:00 AM on the 20th13th.
The phones were clearly not able to re-register to alternate resources. As only some stations deregistered, but all Gate Keepers were affected, the following conclusions can be reached at a glance:
- The incident was not caused by a server or application issue.
- The incident was not caused by the failure of one or more Gate Keepers (say in one Port Network).
- No Gate Keepers had their loading increase, therefore the phones could not see alternatives and connect to them.
Root Cause Analysis at a glance
- The issue was definitely network related.
- It was linked to IP Network subnet(s) on the edge of the network (not core), as some phones remained registered.
Example 3
In the next example below, all most H.323 registrations are currently against only one Gate Keeper even though there are possible alternatives - meaning a single point of failure exists.
The following actions should be taken:
- At least one more Gate Keeper should be configured immediately to remove the single point of failure.
- The enterprise security policy to log out of the phone is not being followed by most users, line managers should be notified.
