mike's blog

We are currently tracking an issue where some inbound calls only last 3-5 seconds and there is one-way audio. The issue is limited to a single carrier. That carrier has been notified and the issue has been escalated.

Outbound calling is unaffected.

We will update this ticket.

Update: 11/1 9:40 AM EDT
We have been checking in with the upstream carrier, and this is still open issue. We will update this ticket with more information.

Update: 11/1 12:10 PM EDT

Update: 11:19 AM ET

Our upstream carrier has corrected the issue. All inbound calls are now completing. We thank you for your patience.

11:04 AM ET

Currently investigating an inbound calling issue. A small subset of inbound numbers are currently unreachable. We have escalated with our upstream carrier and will post updates as we receive them.

Update: 4:28 PM EDT

Service has been restored.

3:59 PM EDT

We are currently experiencing an issue with some of our inbound phone numbers. We have escalated the issue with the affected upstream carrier. We will post updates as we receive them.

Outbound service is unaffected.

Gartner released an interesting report stating that the role of IT should change from one of implementation to one of planning.

With the abundance of cloud-based applications, I can't agree more. The build-vs-buy scenario doesn't even make sense any more. There is such a plethora of cloud-based applications, like the OnSIP Hosted PBX, that running your own in-house service, in most cases, does not make fiscal or strategic sense.

RFO affecting some customers approximately 1:10 - 1:40 EDT, 5/31/11

Some OnSIP customers experienced a loss of service on Tuesday 5/31/11 from approximately 1:10 pm until approximately 1:40pm. This loss of service would have included the inability to initiate an outbound call or to properly answer an inbound call. It was soon determined that the service interruption was do to a UUNet (Verizon) routing issue inside their network between LA and NY. Verizon networks were unable to reliably route traffic to large blocks of IP address, which contained OnSIP IP blocks.

Update: 12:30 PM EDT

All issues are now resolved. This affected 25% of the customer base. Services began restoring at approximately 11:45 AM EDT for some customers. All issues are now resolved. Please feel free to contact Support if you have any further questions.

Update: 11:52 AM EDT

A new DNS entry was put in place at 11:15 to move customers to an alternate NYC proxy. In addition at about the same time the network BGP connection stabilized.

We had a momentary network disruption. The issue has been resolved. As we learn more, we will post.

We recently updated the software (SIP.ld) on the Polycom boot server to version 3.3.1. The new software is more in line with Dynamic Host Configuration Protocol (DHCP) standards.

In the older version of SIP.ld, the program was reading the NTP server from the config file. However, the proper behavior is for the phone to use the NTP server as defined by the DHCP server, regardless of what's in the config file; the config file entry is only supposed to be used if the DHCP server fails to supply an NTP server.

The server sip.onsip.com is currently under heavy load. There is a device on the Internet sending us thousands of bad packets per second which is filling the error logs and bogging down the server. We are attempting to isolate that traffic and allow the server to return to normal. Until that device either stops or we re-route the erroneous traffic, OnSIP service will be sluggish.

Junction Networks is currently tracking an issue where no traffic is able to be routed to any Junction Networks domain. No phones are able to register and no calls are possible either in or out of the system. Engineers have been dispatched. We will update this Network Alert in the comments.