Sunday not shaping up well for Second Life either
The new Second Life day is off to a poor start with what looks like another database crash inside Linden Lab's array of database clusters.
The crash (if that's exactly what caused it) took place at 1:25AM SLT (US Pacific time), one of the quietest parts of the day with only 32,055 users online. Obviously the problem is not solely load-related.
Within several minutes as many as 2,000 users were cut off, or chose to disconnect due to the problem. It's early yet, and this does not bode well for the day. Whatever the cause of the problem is, it's obvious that Linden Lab has not found and fixed it. It seems logical, therefore, that we can expect more of this through the day today - and Sunday is the busiest day of the Second Life week.
Linden Lab has not issued any notice or acknowledgement of this-morning's outage at the present time.
[Update: 7:45AM SLT - A second failure took place from 5:50AM to 6:20AM. Definitely longer than the first one]
[Update: 5:15PM SLT - Two more failures through the day. 10:50AM (6,000 users dropped) and the grid is just recovering from another at 4:05PM (another 6,000 users dropped - or maybe it's the same six thousand, and they're getting very tired of it.) - Still no word from Linden Lab about today's outages, or whether the cause is being worked on.]
[Update: 7:00PM SLT - Statistical feeds have been erratic for the last two hours, having gone completely daark for the last 15 minutes. That's usually a sign of severe backbone congestion.]
[Update: 7:30PM SLT - The feed failure did indeed signal an outage. We caught some of the details as the statistical feeds recovered, but it looks like a duration of approximately 20 minutes, ending at 7:15PM with at least 4,000 affected users]
[Update: 11:30 SLT - Another glitch at 9PM, and again at 10PM. Today has not been a shining day for the platform, especially on what should have been the busiest day of the week. Linden Lab remains silent.]
