Blocked on server side?

NUCCpod_NAPTIMELABS_01
NUCCpod_NAPTIME...
Joined: 18 Aug 17
Posts: 3
Credit: 25569614
RAC: 0
Topic 213547

I was moving some workers (30 or so) over to einstein@home and after a few moments all workers backed off of download tasks and could no longer communicate with project server.

Port 443 (https) seems to be blocked from this IP as I can also not navigate to this site in the browser.

Utilizing a different network everything seems to browse fine, but that still leaves my workers without access to the project server.

Shawn Kwang
Shawn Kwang
Joined: 3 Nov 15
Posts: 289
Credit: 2740270
RAC: 1029

Yes, you were blocked

Yes, you were blocked server-side, but not intentionally! We have a DOS filter setup to prevent against abuse. When you connected that many hosts to our project, that activity managed to trigger the filter. The block only lasts 24-hours, although I'll try to unban your IP manually in a moment. Sorry about this.

Although we try to design these things to minimize false positives, it sometimes happens. This shouldn't be a problem going forward (hopefully). It was probably the act of having all 30 hosts contact the project simultaneously that triggered the DOS filter. There are plenty of other users with dozens hosts so normal operations shouldn't be affected.

Einstein@Home Project

NUCCpod_NAPTIMELABS_01
NUCCpod_NAPTIME...
Joined: 18 Aug 17
Posts: 3
Credit: 25569614
RAC: 0

I figured this was the case,

I figured this was the case, thank you for the quick follow up. Connection seems to have been restored.

 

Moving forward, I see I am going to have to stagger out workers connecting to be a little more friendly!

 

Thanks again!

 

 

mmonnin
mmonnin
Joined: 29 May 16
Posts: 291
Credit: 3229677276
RAC: 1102757

That's a lot of computers. I

That's a lot of computers. I was curious and took a look at what was attached. Mostly GPU tasks were downloaded from what I could tell. Some of the hosts don't have enough vram to run E@H though. Like the GeForce 9800 GT (511MB). :(

 

One of the GeForce GT 630 (963MB) has errors, it might be out of vram esp if there is some used for a DE.

NUCCpod_NAPTIMELABS_01
NUCCpod_NAPTIME...
Joined: 18 Aug 17
Posts: 3
Credit: 25569614
RAC: 0

mmonnin wrote:One of

mmonnin wrote:
One of the GeForce GT 630 (963MB) has errors, it might be out of vram esp if there is some used for a DE.

 

Nice catch, that particular machine is a bit wonky.

I dialed it in a bit just now, I'll keep an eye on it today as it ticks along.

Gary Roberts
Gary Roberts
Moderator
Joined: 9 Feb 05
Posts: 5842
Credit: 109403791314
RAC: 35505030

NUCCpod_NAPTIMELABS_01

NUCCpod_NAPTIMELABS_01 wrote:
... Moving forward, I see I am going to have to stagger out workers connecting to be a little more friendly!

You shouldn't have to do anything, going forward :-).  The odds of more than a couple of machines trying to contact at much the same time, from here on, must be pretty small.  Of course, if you were going to add a further big batch, just introduce a small delay between each one :-).

Anyway, welcome to the project.  I hope you enjoy your time here!

 

Cheers,
Gary.

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.