Ask network admins how they monitor their IP telephony performance, and many will probably offer some variation on, “We just play it by ear. Literally.” But this can be a crucial mistake. Here’s how to know whether (and how) to proactively monitor your IP telephony environment.
IP telephony quickly became the new bully on the network block, enforcing its “me first” attitude on the rest of the data stream. Mostly, that’s been a good thing. IP telephony can significantly reduce overhead, and it has huge potential for integrating communication tools.
But then there are the times when you’re on an important call or teleconference, when intermittent latency or jitter make people step on one another’s sentences or miss key information because the speaker is cutting in and out.
Or worse, in the middle of an urgent conversation, you finish a long explanation about a complex topic when you realize nobody’s on the other end anymore.
These situations can be anything from a minor annoyance to a major threat to your business. You’re especially vulnerable in these circumstances:
- Your phone system is your company’s bread and butter: e.g. phone sales drive your revenue, or you maintain call centers or help desks that are essential for customer satisfaction, employee productivity, or both.
- You outsource your IP telephony system.
- You use multiple vendors for SIP enabled phones, network gear, monitoring software, etc.
If your company is dealing with one or more of the above circumstances, consider these tactics for giving your staff the reliable IP telephony system they need to do their jobs, and delivering the cost savings you should be able to expect:
1. Measure QoS More Than One Way
If you’ve graduated from measuring IP telephony system performance strictly by waiting for problem reports to actively monitoring it, congratulations. But if you’re using only your call manager’s metrics, you need a backup plan.
In addition to polling call detail records from the IP telephony central server, you should be continually measuring the latency rate, jitter, etc. of sample packets exchanged by routers a significant distance apart.
Set alert thresholds for latency, jitter, and packets dropped during these sample data exchanges. This can help you detect potential issues even when your phone traffic is low or non-existent, so you’re prepared to deal a bottleneck or a bad network before it becomes a problem.
Some IP telephony system monitoring tools use one or the other of these measurement methods. We insist on redundancy in other monitoring environments, and IP telephony should be no different.
2. Seek Additional Levels of Visibility Into IP telephony Data
Some IP telephony tools have useful monitoring features that you can enable and configure to provide visibility into that specific tool’s performance. This type of monitoring, however, may not give you an accurate picture of your complete IT telephony environment.
In some cases, unfortunately, product-specific monitoring capabilities sometimes lag product releases by a good several months or even longer. They tend to provide only one basic level of visibility.
That’s one reason many telephony monitoring tools report your call quality based solely on an average of all your calls within a given period. The trouble with this method is that it doesn’t always tell you when a single, critical call channel is entirely blocked.
For example, say your most critical call traffic goes outside your company’s dedicated internal IP telephony network, but 90% of all call traffic is inside that network. The average call quality might be within acceptable parameters, even when your people can’t reach anyone outside the network.
That’s a bad blind spot to have.
3. Monitor IP Telephony as One Part of the Larger Picture
When you’re monitoring your IP telephony’s traffic levels, it helps to have the health of every related piece of hardware — routers, switches, etc. — and anything else in between your data transmissions and their final destinations.
You should be able to see at a glance whether a single part or remote location is a data pinch point, whether it’s originating on- or off-premises.
Another reason to expand your monitoring system’s visibility is to accommodate legacy systems, varying hardware and software vendors, and external hosting services. The individual monitoring processes for these elements need to be integrated.
If they’re not integrated, you’re probably devoting unnecessary staff resources to learning and maintaining the monitoring systems separately. It’s also easier for communications to break down, the more layers of specialization are involved.
4. Take More Control Over Profitability
A great deal of the savings you can expect from IP Telephony is savings for long-distance calls. Don’t be satisfied with any savings — make sure your calls are being routed without unreasonable detours that cost you money.
We all know there’s still plenty of the Wild Wild West in the World Wide Web. Data can be mis-directed for countless reasons, and some data wranglers benefit when that happens with VoIP data. Service level agreements (SLAs) can incent your partners to help you prevent this.
You can seek guarantees for certain performance parameters, such as a WAN provider’s throughput, availability, latency, jitter and packet loss.
Overall, profitability comes down to making the right decisions. To give your company the best chance of making good decisions about its IP telephony platform, your monitoring system should give management a single high-level view that accurately reflects what the engineers are seeing.