January 5, 2018 Platform Update
-
MangoLassi will be going offline for a little over an hour this evening during a system hardware migration. We are upgrading to faster hardware, and beefier hardware. The move is estimated at 85 minutes. Fingers crossed that it will be less. We will let the datacenter know to pop us into the work queue as soon as we get to our dead Friday evening period. We won't be able to predict really well when we will go down, as there might be a queue. So expect to see some emptiness this evening, but we should be back relatively soon.
-
Yay, time for a party.
-
I haven't thought things were running slow.
But of course I don't see the performance metrics like @scottalanmiller does.
-
@dustinb3403 said in January 5, 2018 Platform Update:
I haven't thought things were running slow.
But of course I don't see the performance metrics like @scottalanmiller does.
Nothing he stated implies that ML is running slow.
-
@jaredbusch So what is the motivation to move to a more beefy server for ML then?
-
@dustinb3403 said in January 5, 2018 Platform Update:
@jaredbusch So what is the motivation to move to a more beefy server for ML then?
Datacenter refresh to newer, bigger hardware.
-
@dustinb3403 said in January 5, 2018 Platform Update:
@jaredbusch So what is the motivation to move to a more beefy server for ML then?
Don't know, and don't care.
But your statement has nothing to do with the announcement.
-
@jaredbusch said in January 5, 2018 Platform Update:
@dustinb3403 said in January 5, 2018 Platform Update:
@jaredbusch So what is the motivation to move to a more beefy server for ML then?
Don't know, and don't care.
But your statement has nothing to do with the announcement.
It was more an inquiry than it was a statement. . .
-
@scottalanmiller said in January 5, 2018 Platform Update:
@dustinb3403 said in January 5, 2018 Platform Update:
@jaredbusch So what is the motivation to move to a more beefy server for ML then?
Datacenter refresh to newer, bigger hardware.
Nice, thank you for clarifying.
-
We gain 50% more RAM in the move, too. We don't exactly need it, but we are getting closer and closer to it.
-
Prepping now for the outage.
-
Offsite backups are complete. We are requesting the datacenter queue.
-
Queue entered. We are first in line. Should go down momentarily, I would guess.
-
Estimate is still 85 minutes for the process to complete.
-
And we are back!
-
Took only 18 minutes, not too shabby.
-
Looks the same
-
Ya def feels faster. Especially on Firefox on iOS. I hated using it for this site but it’s really quick now. But that could be because no one is here yet.
-
I think so. Feels faster here, too. It's newer CPUs (but not more cores), 50% more RAM, a few tiny OS updates (we ran updates just before the migration), a fresh reboot.... no one major thing, but several little things all at the same time. So very possibly faster.
-
Working good here
