Due to timezone conversions, five minutes is now five weeks.
Printable View
Due to timezone conversions, five minutes is now five weeks.
It has now been concluded that quantum fluctuations emanating from the server have created a localised 'time dilation bubble' creating inconsistencies in the database, as a result we can expect the tech to reach the server with the crash cart in a time asymptomatically approaching infinite as he gets closer to the server.
Unfortunately Experts at CERN have given us estimates of four weeks before before the whole of the UK have succumbed to the effects of the rapidly expanding bubble. By the time we have launched Bruce Willis in a rocket in order to destroy the server with a thermonuclear device it may already be too late. On the plus side, if he succeeds the simultaneous explosion and implosion will be contained by the time dilation bubble meaning only 1/4 of the Manchester metropolitan area will be levelled.
time dilation bubble?
timey wimey wibbly wobbly sounds better
what a long memtest that was.
You sure do pick the best time to perform maintenance that you know will take hours.
What was the results of the initial memory test? Did it provide results that would indicate there's faults in the memory hence why you decided to perform a longer memory test? Have you replaced memory due to faults?
The announcement summed up is 'we performed short memory test to see if it has any issues, we then performed a full memory test, as you can see the forum is stable'.. well the forum appeared to be stable the rest of the time, apart from when it was dead so not really helping but yeah
Which databases were corrupt? Was there any data loss? Was the reboots down to issue with RAM? Has this now been fixed? How many times did the reboots corrupt databases before it was fixed? How do you plan on avoiding this in future if there is similar RAM issues?
They need to wait for Jin for the results of the test. They started the tests at 9AM because that was the earliest convenient time and Recursion has already spent enough time rebuilding databases to wait longer.
As for which databases were corrupt, from the ones Recursion has named I can't think of any which weren't corrupt, there may have been some recoverable data loss. I believe there were two unexpected restarts/shutdowns and they corrupted another time in the middle of fixing one of them. I suspect similar RAM issues are unlikely (If there even are RAM issues, which there probably aren't since it's ECC).
How come you provide us with more information than any actual staff member lol..
There must have been some indication of RAM issue if they prompted for a full memory test though, otherwise doing a short one followed by a full one was just stupid when you could go for the full one first and take a lot less time.
Was there any unrecoverable data loss?
Will hasn't really provided much extra information then what is already out there by either me or Tom. We also stated the following:
Considering this isn't over then a lot of the information hasn't been published at this moment in time. Not like the majority of the forum userbase would understand it either if it was just full of technical terms.Quote:
We shall also post more information about the recent issues once we have all of the Habbox websites back up and running.
I will so post it
Yeah sure, later.Quote:
Originally Posted by scottish