Starting at approximately 5:45am SLT (US Pacific), we began to receive reports consistent with backbone problems. These include (but are not limited to) login problems, asset and inventory problems, teleportation problems, search problems, and difficulties with loading website pages (particularly F1/Help, but also all of the access-controlled portions of the site).User concurrency at the start of the difficulties was 28,150. While the problems do not seem to be severe enough to cause mass logouts, the rise in concurrent users has been slower than usual for this part of the day. At present the issues seem to be coming in short waves and are nowhere nearly as severe as many recent problems.
A little more than five hours later there is no word yet from Linden Lab about the cause of the difficulties or the prognosis for the rest of the day.
[Update: Linden Lab posted about this, about a quarter of an hour after we did. The last informational update (at approximately 8pm SLT (US Pacific) was that the problems seemed to be easing (and indeed they have!) but that work on it was still continuing. Cyn promises details once the problem is resolved.]
An interesting note, the backbone problems ceased around about the time that the concurrency dropped back below 28,000 or so. We don't think that means it has been fixed, but that there is a load-related issue going on behind the scenes.














1. As of 11:16 SLT, the Grand Unified Linden Blog is reporting that the asset server problems are back. It may be that that issue is not quite as "hardware related" as LL thought it was yesterday, or it may be that the machines they shifted the load to are just as creaky as the ones that were failing yesterday. It may also be that several go-rounds may be needed to finally nail this particular problem to the wall.
Posted at 2:44PM on Jun 14th 2007 by Erbo Evans