Wed Apr 15 01:37:21 PDT 1998 — Our T3 to UUNet is currently offline due to PacBell maintenance. Pacific Bell estimates that they will be finished in about an hour, and says that they did not anticipate taking any customers offline. As our two T1 circuits are very lightly loaded at this time of night, you should not notice any impact at all due to this. Redundancy is very important in an ISP – remember that when you’re shopping around! -Dane
A power event a few minutes ago made a reboot
Tue Apr 14 15:03:38 PDT 1998 — A power event a few minutes ago made a reboot necessary on our Pop Mail, and List delevery hosts. The event was interesting in that it bypassed our backup power supplies! Total downtime was less than 15 minutes, during which Email services were not available. Sorry for the delay! -Eli
Update – it wasn’t a power event, it was a network attack. Someone sent an improperly formed packet to a huge number of systems on our network, and some which had not been updated crashed and rebooted themselves. Affected systems were tsunami (add-on mailbox accounts), afterburner (list processing), marine (outbound mail relay), blam and slam (local routers) and a few tech workstations. These systems had not been updated to Linux kernel 2.0.32+. Shouldn’t happen again, they’ve all got new kernels! -Dane
UUNet has been experiencing routing…
Thu Apr 9 10:38:12 PDT 1998 — UUNet has been experiencing routing instability since about 8am this morning. This can cause connectivity problems with some networks on the Internet. We’re working to get an ETR, and evaluating what we can do to minimize impact to our customers. (Fortunately, our Sprint and MCI connections are fine. 🙂 -Scott
The equipment which serves 522-1003 had…
Tue Mar 31 21:21:05 PST 1998 — The equipment which serves 522-1003 had trouble (again) this evening and had to be rebooted. This equipment serves overflow for all of the other numbers, so if you had your connection terminated once at about 9:15 this evening, or experienced busy signals between 8:50 and 9:15, that was probably the trouble. We’re working with USR, and are swapping one component tonight at 4am to try to isolate the problem. -Dane and Scott
We’re currently testing a new modem code…
Tue Mar 31 14:46:43 PST 1998 — We’re currently testing a new modem code release which includes support for the new v.90 56k standard. If you’ve got a v.90 enabled modem and would like to give it a try, you can dial 707-522-1389 to reach the test lines. If you do try it and have feedback about how it works for you, please email support@sonic.net. -Brian and Dane
Our web server was running at a greatly…
Fri Mar 27 04:53:03 PST 1998 — Our web server was running at a greatly diminished capacity between 3:50AM and 4:35AM this morning. It appears to have run out of RAM and was swap bound. The system has 256 megs of ram and is typicly very lightly loaded at this time of night, so we’re not sure why this would happen. We’re investigating further in the logs.
Our ‘who’s online’ web page is back in action
Tue Mar 24 23:04:46 PST 1998 — Our ‘who’s online’ web page is back in action after being offline because of a RADIUS upgrade. You can view it from the list at www.sonic.net/info/. -Dane
PacBell recently installed 96 new lines on…
Tue Mar 24 21:18:23 PST 1998 — PacBell recently installed 96 new lines on four T1 PRI trunks, and it seems that they failed to setup the hunting so that they would be usable. We’re opening a repair ticket with PacBell right now, and expect to have this fixed before prime time tommorow. Sorry for the busy signals tonight. Note that once this is correctly configured, we will have 887 total ports. -Dane
Our line utilization graph is back online…
Thu Mar 19 09:52:10 PST 1998 — Our line utilization graph is back online after a month long absense. It was offline because we upgraded the RADIUS authentication server to a faster better version, and the new version didn’t have the features to track usage this way. We’ve changed the way that we’re gathering this data, and it’s back online. The line utilization graph is again available at: www.sonic.net/lines/
We’ve changed the way that the FTP space is…
Wed Mar 18 10:02:50 PST 1998 — We’ve changed the way that the FTP space is mounted on the server so that the .snapshot live backup filesystem is not visible within the public ftp space. We’re hoping that this will solve problems with customers ending up in the snapshot area where they cannot upload, and with remote FTP users getting files which are out of date.