[DOWN] Bitlc.net - P2SH, No invalid blocks, Custom payouts, EU, 0% fee, LP

[crazy987]

i cannot connect from NL

[1713957688]

1713957688

[1713957688]

1713957688

[Jine]

THIS JUST IN:

Jine is in DC with his girlfriend eating cheeseburgers!

No I'm not? I was asleep. Stop spreading bullshit, it's enough with your trolling in our irc-channel.

The issue currently is that I'm seeing huge amounts of getworks from a proxy, that makes one node hang.
When pfSense detects a node non-functional, it shuts it down and transfers the connections to another node.

... which then creates a spiral where all nodes go *poff* one at the time. After a few minutes, those other nodes becomes responsive again - and they are back online. After a while, the proxy/whatever is back on that same node again, and makes it hung once more.

I banned the proxy a few seconds ago, when i woke up. (Or to be correct, i banned a few subnets, to avoid the "bad guy" just switching VPS/IP/etc.)
I'm sorry for the instability during the night. I'm currently trying to setup a traffic shaper/limiter for amount of connections.

[Dusty]

I want to simply thank Jine for his dedication on this (difficult) task. Thanks!

And long live bitcoins.lc

[Jine]

I want to simply thank Jine for his dedication on this (difficult) task. Thanks!

And long live bitcoins.lc

Thanks!

And if anyone have suggestions on what we should do to avoid this in the future, please feel free to help out.
I'm starting to get out of ideas.

[Jack of Diamonds]

The issue currently is that I'm seeing huge amounts of getworks from a proxy, that makes one node hang.
When pfSense detects a node non-functional, it shuts it down and transfers the connections to another node.

... which then creates a spiral where all nodes go *poff* one at the time. After a few minutes, those other nodes becomes responsive again - and they are back online. After a while, the proxy/whatever is back on that same node again, and makes it hung once more.

Substantial amount of getworks per second & the use of VPN/proxy suggests an inefficient, large CPU miner network (very likely botnet),
probably selecting this pool for maximum efficiency due to 0% fees.

Maybe a script can be made which restricts getworks based on total valid shares being submitted from that IP range
(because bots will only be sending a few shares per hour due to a majority of infected computers being CPU's only, leading to a very inefficient workload which stresses the server)

[Dusty]

And if anyone have suggestions on what we should do to avoid this in the future, please feel free to help out.
I'm starting to get out of ideas.

Jack's idea seems to me the most interesting one: instead of just restricting number of getWorks (or, more in general, web access) to an IP you should use the worker name as a discriminator.

In other words there must be a minimum ratio between number of web requests and number of valid shares per unit time. This way a large (and productive) user behind a proxy is not penalized while a bot network issuing requests from all over the world (and using the same worker name) can be correctly detected and shutted down.

I don't know php but I would handle this kind of policy using a filter under tomcat (a java app server), this way it's not handled by the application but it's simply an access layer.

[Sukrim]

And if anyone have suggestions on what we should do to avoid this in the future, please feel free to help out.
I'm starting to get out of ideas.

Jack's idea seems to me the most interesting one: instead of just restricting number of getWorks (or, more in general, web access) to an IP you should use the worker name as a discriminator.

Next step for botnet operator: write a small script to generate 100000 worker names and set the worker name in your trojan payload to random(100000)

[gentakin]

If the last-15-minutes-efficiency (submitted shares per getwork request) on a connection/account drops below x (say, 0.2): limit that connection to max. 1 getwork request per y seconds.

However, this will hurt "backup miners" when their backup pool is bitcoins.lc.. And I'm not sure if it helps at all. Just an idea.

[darkskypoet]

Any way to have a certain min MH/s to play? I mean if the worry is CPU miner botnets, then could you not have one node that handles first attempts / requests, and the other nodes only get involved to serve work after the first node is satisfied that its one of us (a GPU miner). Then a handoff occurs, as we are vetted? So a pool of load balancing nodes once you're in the club so to speak, that way the work is getting done, the majority of us are unaffected, however new requests, new addresses, etc (some way to filter it) can't get past the initial node w/o x share / x time?


It means issues would threaten that initial node, may make it hard to get accepted to the other nodes, but the other nodes would still cheerfully toss out work to the majority of users mining away.

[PabloW]

Sorry to say this Jine, but as you may notice many people are leaving, and thats because many of the promised things, as little as they can be, are not being done. Something basic like auto-payout cant be missing out at this time.

[Dusty]

Next step for botnet operator: write a small script to generate 100000 worker names and set the worker name in your trojan payload to random(100000)

You can simply avoid such a behavior with a captcha (both for account creation and new worker.

Anyway, the more work a bot owner has to do, the more it looks for simpler alternatives, i.e.: less managed pools.

[Jine]

Something basic like auto-payout cant be missing out at this time.

I have multiple times said that that is not going to be implemented due to future plans of the site.
And I'm working as fast as i can with the new system, for each time someone attacks, points botnets or using proxys with multiple hundrets of clients behind it - i have to stop and solve the issues.

Please bare with us. I'll implement stuff as soon as i can

[speckman]

Sorry to say this Jine, but as you may notice many people are leaving, and thats because many of the promised things, as little as they can be, are not being done. Something basic like auto-payout cant be missing out at this time.

No.  Way more likely they're leaving because they can't get work/connections.  I like the payout system the way it is.  I only left because my miners were idle.

And Chick, wtf is wrong with your signature?  Is that your face?  Why would anyone want to see your face x6 every time you say something?  Don't say anything cause I don't want to see it again, but seriously, that's the ugliest, most annoying sig on the site. 

[Chick]

Sorry to say this Jine, but as you may notice many people are leaving, and thats because many of the promised things, as little as they can be, are not being done. Something basic like auto-payout cant be missing out at this time.

No.  Way more likely they're leaving because they can't get work/connections.  I like the payout system the way it is.  I only left because my miners were idle.

And Chick, wtf is wrong with your signature?  Is that your face?  Why would anyone want to see your face x6 every time you say something?  Don't say anything cause I don't want to see it again, but seriously, that's the ugliest, most annoying sig on the site.  

I would take that as a compliment, he's the COO of Nintendo of America. It is quite a face and goes well with 'GOXED'.

[Chick]

THIS JUST IN:

Jine is in DC with his girlfriend eating cheeseburgers!

No I'm not? I was asleep. Stop spreading bullshit, it's enough with your trolling in our irc-channel.

The issue currently is that I'm seeing huge amounts of getworks from a proxy, that makes one node hang.
When pfSense detects a node non-functional, it shuts it down and transfers the connections to another node.

... which then creates a spiral where all nodes go *poff* one at the time. After a few minutes, those other nodes becomes responsive again - and they are back online. After a while, the proxy/whatever is back on that same node again, and makes it hung once more.

I banned the proxy a few seconds ago, when i woke up. (Or to be correct, i banned a few subnets, to avoid the "bad guy" just switching VPS/IP/etc.)
I'm sorry for the instability during the night. I'm currently trying to setup a traffic shaper/limiter for amount of connections.

I guess you were trolling me. It's quite a feat accomplishing that. 

[speckman]

I would take that as a compliment, he's the COO of Nintendo of America. It is quite a face and goes well with 'GOXED'.

Yeah, I'm not surprised that you'd take that as a compliment because I'm surely not the first person to say Jesus Christ, do we really all have to stare at the face x6 every time you open your mouth?  And I understand that there are all types of people, and one of them is the type that loves to annoy other people.  :|

[V2-V3]

I would take that as a compliment, he's the COO of Nintendo of America. It is quite a face and goes well with 'GOXED'.

Yeah, I'm not surprised that you'd take that as a compliment because I'm surely not the first person to say Jesus Christ, do we really all have to stare at the face x6 every time you open your mouth?  And I understand that there are all types of people, and one of them is the type that loves to annoy other people.  :|

Would you like it better if he used one large image so you only have to see the face once?



Also, nice pool Jine! Keep up the good work

WTF.....

[Jine]

Seriously, please stop spamming this thread with unnecessary posts.

[padrino]

Very impressive! Running for 90 minutes or so and keeping rejects at 0.590%

[pandemic]

Can't mine tonight, tomorrow, or tuesday. We have guests... and it's friggin hot in my apartment right now. Looking forward to winter