(The following was posted to a thread for the issue for Circle PCs being unable to connect and status going red because of it from 8/24 through 8/30/07)
-steve
-------------------------------
Hi everyone,
We've been working hard here to try and understand the issue, and were finally able to track down (what we believe to have been) the root cause.
To provide a high level summary--it essentially seems to have been caused by a couple of unexpected server-side issues working in conjunction, as Steve and others have mentioned. We had a power outage here over the weekend in the buildings hosting the beta servers. During the reinitialization when the machines came back up, there appears to have been a couple of failures, eventually leading to some of the OneCare Circle related server-side functionality being intermittently down. We have a pretty good monitoring in place to alert us to such failures, but unfortunately in this instance there appears to have been a different (unrelated) issue at the same time, that led to these failures not getting reported as we'd have expected.
We think we've now fixed the issues at hand, and the clients should auto-fix themselves in the next few hours. I should note that we've had a couple of learnings from this--both in terms of the actual code bug and monitoring we had in place, but also providing more status as we were debugging the reported failures.
Please let me know if you're still running into issues. My apologies for the issues we encountered, and as always, thank you for trying out the Beta--we really do appreciate it very much!
Thanks,
Guhan
