Fri Jan 11 14:26:05 PST 2002 — The issue with the 9811 pop has been resolved and the pop should now be functioning properly. We’ll keep an eye on it for the next few days. -Kelsey and Russ.

Fri Jan 11 13:14:41 PST 2002 — Users have been reporting problems with our XXX-XXX-9811 pop. We have a trouble ticket open Focal who provides us with the circuits for these numbers and are awaiting a call back from their support department. We do not have an ETR at this time. Users that are experiencing difficulties with a 9811 number may be able to replace the final four digits with ‘9606’ which is our new San Francisco pop. Please note that the 9606 numbers came with community.net and will not work for approximately 300 sonic.net users, who have name conflicts with community.net users, until a later date. Whenever changing the number that you are using to connect to us, please verify that it is a local number to avoid any long distance charges. -Kelsey, Russ and Eli.

Wed Jan 9 11:29:54 PST 2002 — We have been experiencing capacity problems in the Central Valley area. We are aware of the issue and are working on getting more capacity ordered. We will update you with an ETA. In the mean time we will look at ways to alleviate the impact on the POP. – Matt

Mon Jan 7 14:11:38 PST 2002 — The server which handles the backend for the downtown wireless network is offline with a hardware failure. We do not currently have an ETR, but are working with our vendors to get it back up and running as soon as possible. Until we are able to restore the server or bring up another one in it’s place, our downtown wireless network is off-line. -Kelsey and Matt

Update Jan 18 09:23:00 PST 2002: A backup server was brought online while we await a replacement from VALinux.

Fri Jan 4 23:14:02 PST 2002 — Update on the Sebastopol POP, After reseting some ports on the PRIs in the Sebastopol pop, PacBell has cleared the ‘All circuits are busy’ error. We will be keeping an eye on this over the weekend, if the problem continues we will escalate it with in PacBell. – Steve

Fri Jan 4 18:00:16 PST 2002 — Our Sebastopol POP (823-8812) has been returning ‘all circuits busy’ messages intermittently, and we’ve (finally) got confirmation from Pacific Bell that they have identified the issue on their end, and that they are working to repair it. Pacific Bell has not given us an ETR, but we hope to have service fully restored shortly. For alternate dial-up numbers in your area, please check www.sonic.net/cgi-bin/pops.pl – Steve, Eli

Tue Jan 1 20:25:35 PST 2002 — Authentication Server Failure: At 7:01PM both of our radius authentication servers failed. These two servers handle all of the authentication requests for all authenticated services at sonic. After some hurried investigation we found that the static hash table used in the radius server software was being overflowed and and preventing the server from (re)starting properly. Shortly after 8:00PM we had the servers back up and running properly with a properly sized hash. -Kelsey, Nathan, Matt and Russ.

Mon Dec 31 12:53:33 PST 2001 — On December 24th at 3:15pm, freezer, one of our Network Appliance F740 filers failed and icebox, it’s cluster partner, took over its operations. There was no interruption in services during the takeover. This is the first time that we have had an unplanned NetApp cluster failover and it worked flawlessly. Freezer was brought back into service on the 27th after completing diagnostics on the hardware, again with little or no service interruption. We are very pleased with the performance of the NetApp filers and the fact that an otherwise devastating hardware failure went unnoticed by our customers. We were also pleased that we were able to put off fixing the problem until after we had returned from the holidays! -Kelsey, Nathan, Eli and Dane

Mon Dec 31 13:14:37 PST 2001 — Sundry Updates and Changes: We have improved the DOS detection scheme in our POP3 servers to eliminate false positives. In the past, some users may have received an error, “Slow down; you are reconnecting too fast” This primarily effected users behind office firewalls or NAT boxes. If you continue to receive this error, please contact support.

As stated in an earlier MOTD we have migrated archived weblogs to a new directory structure. Archived logs used to be available on the shell and ftp servers at /var/log/httpd/domain.name and are now stored at /var/log/httpd/username/domain.name This completes the final step in the release of our new quota and bandwidth tools and provides extra security since only the owner of an account is able to view the logs for that account. Weblogs for your personal webspace are also archived here as well.

We have worked up a fix for the long standing ‘wavyworld’ problem. User’s personal webspace (sonic.net/~username) is accessible under any domain hosted at sonic.net. Starting Wednesday, access to a ‘~’ URL under a multihomed or parked domain will be restricted to the owner of the domain. Exceptions to this rule can be made by request of the domain owner. Please email support to have exceptions added.

-Kelsey and Nathan

Sat Dec 22 19:46:19 PST 2001 — Microsoft Windows security issue update. Sonic.net has placed filters online which block traffic to ports 1900 and 5000, which the FBI is currently recommending to limit the scope of the current Windows XP security flaw.

As these ports are used for regular traffic as well, this will cause some random failures to establish network connections. We estimate that this will happen in less than one tenth of one percent of attempts. Customers with static IP addresses who need to be exempted from these filters can write to support@sonic.net. -Dane, Eli, Nathan, Kelsey and Russ