Month: June 2001

Sat Jun 30 13:14:13 PDT 2001 — BroadLink seems to have had a brief outage this morning. Sonic.net support received a number of calls from customers who were offline, but they all came back online a few minutes later. If you find that you are offline, please unplug the power to the BroadLink CPE, wait one minute and plug it back in. After it boots (which takes a minute or two), if you still have trouble, please give us a call and we’ll troubleshoot. -Dane and John

Sat Jun 30 13:08:45 PDT 2001 — We’ve seen some transient problems in our Cybercash credit card transaction processing through the end of the week. Late Friday night and early Saturday morning it was having timeouts and errors. If you’re attempting to renew or add a service to your account and you have trouble, please wait a few hours and try again. -Dane

Thu Jun 28 18:44:13 PDT 2001 — Problems with twig. Earlier today, twig, our web-based mail reader, had trouble communicating with mailbox accounts. The trouble was traced to the imap server on Tsunami, which was holding on to vestiges of an NIS configuration that it was set up with yesterday. Setting the NIS domain name to “” (null) fixed the problem. This only affected twig — mailbox pop and smtp were not affected. -Scott, Dane, Kelsey, Matt, Russ, Steve

Wed Jun 27 15:45:11 PDT 2001 — Urchin and web log processing update. Urchin, our web traffic statistic package for multihome customers, stopped functioning on June 8th. We were able to get it back up and running on the 20th. We are aware that the version of Urchin that we are using has a number of bugs and we are currently negotiating with our vendor for an acceptable upgrade path. We are planning a complete overhaul of our web log processing tools over the next few weeks and already have new hardware for the project on site. -Kelsey

Mon Jun 25 09:47:19 PDT 2001 — The 522-1001 dialup group was having issue authenticating this morning. This required a reboot of the gear. Once we rebooted the gear all started working again. Sorry for any problems this may have caused. -Steve

Correction: The dialup group affected was 522-1003 -Steve

Sat Jun 23 02:09:26 PDT 2001 — Redundancy work completed with no downtime. The second switch (Giga) for our fault-tolerant network design is online. We have activated redundant gigabit links from both Network Appliances to both core switches. Not only will this improve performance to the filers, but will ensure that the filers remain available even if a switch fails. A test of the failover capability worked flawlessly.

We have also moved one of our Cisco 7206 border routers to Giga. Both border routers are already injecting redundant default routes into our OSPF routing domain, which means if one border router fails, we won’t lose the Internet.

Kermit, our second Alteon AD3 load-balancing switch, is now alive and well on Giga. We will use Kermit to build out a new active-active load balancing configuration, which we will then synchronize with Piggy, our other load-balancing switch. This will ensure that web and mail remain up even if a switch fails.

Now that we’ve worked around and solved all of our problems with the Black Diamond switch (knock on wood), our redundant design will finally achieve fruition. -Scott and Kelsey

Fri Jun 22 06:56:07 PDT 2001 — Timber’s maintenance was completed as scheduled. Everything looks good. Total downtime for the SQL database was approx. 45 minutes. -Steve

Fri Jun 22 00:36:42 PDT 2001 — Redundancy work. The second switch for our fault-tolerant network design is online. At 1am tomorrow (Saturday) morning, we will be turning up redundant gigabit links from both Network Appliances to both core switches. Not only will this improve performance to the filers, but will ensure that the filers remain available even if a switch fails. After this, we will move one of our Cisco 7206 border routers to the second switch. Both border routers are already injecting redundant default routes into our OSPF routing domain, which means if one border router fails, we won’t lose the Internet. We don’t anticipate any of this work to cause downtime, but we thought we’d better mention it in case something happens. -Scott and Kelsey

Thu Jun 21 16:18:47 PDT 2001 — We will be performing disk maintenance on timber.sonic.net. This is our customer MySQL server. SQL access will be unavailable for approximately one hour starting at 6:00am on June 22, 2001 and will be back up at approximately 7:00am -Steve

Thu Jun 21 15:57:50 PDT 2001 — Redundancy work. The second switch for our fault-tolerant network design is online. At 1am tomorrow (Friday) morning, we will be turning up redundant gigabit links from both Network Appliances to both core switches. Not only will this improve performance to the filers, but will ensure that the filers remain available even if a switch fails. After this, we will move one of our Cisco 7206 border routers to the second switch. Both border routers are already injecting redundant default routes into our OSPF routing domain, which means if one border router fails, we won’t lose the Internet. We don’t anticipate any of this work to cause downtime, but we thought we’d better mention it in case something happens. -Scott and Kelsey