Service Down ?

[VOIPoTim]

Both TCP and UDP? I thought VOIPo suggested that range for UDP only. At least that's what I've seen many times in these forums, and how I have my router configured.

I can't imagine forwarding all those for TCP...

-Craig

UDP only is fine. All our communication is on UDP.

[man]

UDP only is fine. All our communication is on UDP.

That 949 local area code must be helping with these new accounts

Growth is always a good problem to have and many users understand as scaling hick ups gets ironed out.

The maintenance notices on the forum is helpful.
I did notice the forum stays up even when the Sql server is taken down for maint this morning.

Thanks Tim for keeping us in the loop of your hardware upgrades.

[VOIPoTim]

That 949 local area code must be helping with these new accounts

Growth is always a good problem to have and many users understand as scaling hick ups gets ironed out.

The maintenance notices on the forum is helpful.
I did notice the forum stays up even when the Sql server is taken down for maint this morning.

Thanks Tim for keeping us in the loop of your hardware upgrades.

Thanks for the comment. We do try to always keep lines of communication open. The forums are on a different database server so they'll be up during any issues or maintenance. The helpdesk is also hosted in a completely different datacenter entire so that it always remains up for support submissions during outages or issues.

[zcnkac]

I have modified my asterisk to measure how long it takes to register with the SIP server (usually central01). Over the past two weeks, I have seen spikes of up to 15 seconds that happen at least once a day. Even today after the database server maintenance, I had a spike of 14 seconds to west01. I have had better luck with west01 excepting today, with a spike of 6 seconds.

I experienced a 'not in service' message once. As has been previously mentioned, a delayed database lookup can cause this and also latency in registrations.

[VOIPoTim]

I have modified my asterisk to measure how long it takes to register with the SIP server (usually central01). Over the past two weeks, I have seen spikes of up to 15 seconds that happen at least once a day. Even today after the database server maintenance, I had a spike of 14 seconds to west01. I have had better luck with west01 excepting today, with a spike of 6 seconds.

I experienced a 'not in service' message once. As has been previously mentioned, a delayed database lookup can cause this and also latency in registrations.

Yeah it definitely could. 15 seconds is still within the industry standard for registrations though, so I wouldn't worry about that much. It's a lot more tolerant.

Incoming calls are different since a 3rd party will set their own timeouts before they consider failure.

We'll continue to keep an eye on things to see if there are any new load spikes like the handful we saw causing failed calls.

[bakntime]

Thought I'd pop back in and say that they quickly fixed my problem (outbound call to specific number was reporting "not in service" message). Customer service has always been quick and responsive, and in a still-emerging technology like voip there are going to be the occasional flukes that you wouldn't have had to deal with on POTS. But then again, paying $10/month vs. paying $70/month isn't much of a contest. There are bound to be a few growing pains, but at least we know that Tim and the support team aren't resting on their laurels - they're active on the forum and support is easy to contact via ticket, e-mail, and phone. It's good to know they care, and I've always felt confident that if there was an issue, they'd be right on it.

Aside from the occasional blips over the past 8 months, service is relaible and quality is great.