Month: May 2003

Name server issues.

Fri May 30 09:57:32 PDT 2003 — Name server issues. We’ve had occasional issues with our nameservers, which can cause slow queries as the queries fail over to redundant nameservers. We think that doubling the 1GB of memory to 2GB will greatly improve the situation, and toward that end we have three 2GB servers that will shortly become our new nameserver cluster. -Scott, Kelsey, and Operations

Mail Server Migration: On Monday the 2nd of…

Tue May 27 10:08:52 PDT 2003 — Mail Server Migration: On Monday the 2nd of June we will be migrating ‘’ to a new cluster of servers. This is the first stage in the complete re-architecture of our mail server infrastructure. There are a number of significant changes on the new servers which may affect some customers. We have replaced POPB4SMTP authentication with SMTPAUTH and enabled STARTTLS. SMTPAUTH and STARTTLS are well defined standards that allow users to securely authenticate and relay mail off our servers while they are not directly connected to our network. Some users, connecting from remote networks may need to make a few minor configuration changes to their email clients. If you have any questions please contact Technical Support or post them to news:// -Kelsey

Our free DSL equipment (with rebate!) offer…

Tue May 27 13:17:19 PDT 2003 — Our free DSL equipment (with rebate!) offer has been extended by 24 hours, to this Thursday the 29th at Midnight. We are trying to hit a 5% growth target (or SBC doesn’t fund it, leaving us absorbing the costs), so if you were thinking of switching to DSL, now would be a great time, at least from our perspective. =) -Dane

NetApp filer lockup.

Mon May 25 10:30:17 PDT 2003 — NetApp filer lockup. One of our NetApp filers locked up this evening causing the redundant filer to perform an automated take-over. After investigation, no cause for the failure could be determined. We rebooted the troubled filer and transfered control back to the unit and both filer are functioning normally now. has invested in a redundant, fail-safe storage system which is designed to handle such events. The failure of a filer is transparent to customers as when one filer experiences problems which prevent data from being served, the other file takes over control of the downed filers disks and serves its data seamlessly. Our Network Appliance F740 network file system cluster is the basis of’s storage solution. All user data resides on the two NetApp filers, and they’re configured in a completely redundant configuration. The drives are dual-channel fiber arbitrated loops, and are served by redundant processor heads. The units are RAID level four, WAFL filesystems, and include redundant power and cabling to all disks and the network itself at Gigabit speeds. -Matt and Kelsey

San Jose DSL connectivity issues.

Thu May 22 08:52:45 PDT 2003 — San Jose DSL connectivity issues. A handful of San Jose ADSL customer experienced a brief loss of connectivity starting at 9:30am. The outage lasted about 10 minutes. The problem has now cleared and we are working ASI to determine the cause. -Matt and Tech Support

Graymail Updates: In order to control…

Mon May 19 12:15:30 PDT 2003 — Graymail Updates: In order to control resource consumption and to increase the effectiveness of the web based Graymail interface we have instituted Graymail quotas. We have set a default quota for the maximum number of messages that can be left in Graymail to 1000. The graymail processor will remove old messages in excess of this quota. Users can change this number, as well as the number of days Graymail is kept in the Graymail preferences. We have also confirmed some reports that users are not receiving the notifications and are currently investigating. -Kelsey and Nathan support will be closing one hour…

Fri May 16 11:28:54 PDT 2003 — support will be closing one hour early on Saturday the 17th, at 9:00PM instead of our usual 10:00PM closing. We’re taking the staff and guests to see The Matrix Reloaded. Thanks go to the Roxy Theater for helping us put this together. -Dane, Jen and Eli

Broadlink tower outage – The Wally Moore…

Wed May 14 12:27:34 PDT 2003 — Broadlink tower outage – The Wally Moore tower that serves southern Santa Rosa is unreachable, and Broadlink has dispatched an engineer to the tower. No estimate on a repair time yet, watch this space for updates. – Eli, Jason/BL Update 13:04hrs – PG&E service to the tower site failed, and BL is en route to the tower with a generator to restore service. We expect the tower to come up in 20-30 minutes. – Eli, Broadlink

The Tower came back online at 13:30 as planned.

Night Operations: Thursday morning at 1:00AM…

Tue May 13 15:42:29 PDT 2003 — Night Operations: Thursday morning at 1:00AM we will be upgrading, one of our two Foundry Networks switches that provides redundant network services to colos, to it’s latest stable code release. The new release provides a number of bug fixes and some new features. This upgrade should not interrupt any services to our customers. After the new code has proven to be stable, we will upgrade to the same code. -Kelsey and Nathan

Update: We have postponed this night ops until next week. -Kelsey