search

Tribler / dispersy

The elastic database system. A database designed for P2P-like scenarios, where potentially millions of computers send database updates around.
http://www.tribler.org
87 stars 43 forks source link

We are not power cycling the NAT boxes between experiments #2

Closed whirm closed 11 years ago

whirm commented 11 years ago

We should get a usb/serial controlled power strip to be able to automate power cycling all the NAT boxes before or after a NAT experiment run.

Something similar to this would be good enough: http://www.amazon.com/PwrUSB-Basic-controlled-Automation-automatically-remotely/dp/B008AZW9P6

whirm commented 11 years ago

Power strip ordered

whirm commented 11 years ago

I'm getting the ticket again as the power strip is been ordered.

whirm commented 11 years ago

The power strip has been installed and we have a script in node300 that allows us to power-cycle all the nat boxes.

We need to decide when to do it.

boudewijn-tribler commented 11 years ago

When do the tasks run that use the NAT boxes? Would x minutes before those tasks start be doable in Jenkins?

whirm commented 11 years ago

the thing is that more people is using the cluster, so maybe its better to just do it once during the weekend?

boudewijn-tribler commented 11 years ago

I don't know how long it takes before the boxes need a hard reset, if it takes longer than a week, then at the start of the weekend makes sense. I see no problem as long as everyone using the cluster knows about it.

whirm commented 11 years ago

I can't advance on this one until we know what happens with the whole DAS2 thing.

whirm commented 11 years ago

Forgot to close this one. Now the NAT boxes get power cycled before the nightly run.