Bit Super Lab

[cgafeng]

good job and thinks for you share.

[bitmeat]

awesome .. you give us some source code or at least a rough sketch on how you parse the blockchain?

Importer was done in Python, but I don't think the Blockchain will be enough for all the good stats.

Code: [Select]

def call(method, params):
  url = "http://SERVERIP:PORT/rpc"
  auth = (<RPCUSER>, <RPCPASS>)
  headers = {'content-type': 'application/json'}
  payload = {
    "method": method,
    "params": params,
    "jsonrpc": "2.0",
    "id": 1
  }
  return requests.post(url, data=json.dumps(payload), headers=headers, auth=auth)

Make sure to set SERVERIP, PORT, RPCUSER, RPCPASS in the above example. Then you can do things like:

Code: [Select]

response = call("blockchain_get_blockcount", [])
response = call("blockchain_get_block", [x])
response = call("blockchain_get_block_signee", [x])
response = call("blockchain_list_missing_block_delegates", [x])

Please share ideas for improvements, donate to support the effort (I'm running 3 servers costing $100/mo + putting my time on the weekends), community support would keep me interested.

Cheers!

[xeroc]

awesome .. you give us some source code or at least a rough sketch on how you parse the blockchain?

[bitmeat]

I've started on making some basic stats visible, the network was sufferring at some point, in fast it was the most stable when the initXX delegates were live in the beginning it was doing 100%. That hasn't happened since.

I will try to make the stats publicly available live this weekend.

Cheers!

(See my signature for my progress)

[Gentso1]

I am posting here because as far as I know I am the first delegate to take advantage of bitsuperlabs service. I can tell you he and his team have been great. I have asked lots of questions and they come back with answers very quickly. So if you support bitsuperlabs please support


 bitsuperlab.gentso     

[Pheonike]

I would like to participate. I am a network admin and in the process of setting some vps. My linux skills need some work, but this a project i am willing develop them with.

[JoeyD]

If security, ddos and uptime are concerns then we also need info on how the network detects and connects to delegates or if the model is reversed how delegates connect to the network.  That way we might also be able to setup failover systems, traffic shaping and firewalls. Unfortunately I can't deduce this from the source-code because I can't code and I probably won't understand the code if I read it. (Yes I'm mentally disabled, but there's not much I can do about that.)

PS
Seems the need for bitYuan is becoming dire these last few days, so lets hope we can get the show on the road quickly enough.

[svk]

I'm not interested in running a delegate as a profession or company, but I'd be willing to help bootstrap the system and setup a delegate in the meantime, until the system has grown big enough.

I'd have to setup a cloudbased VPS for that though, because my home internet connection is rubbish and there's isn't much I can do to improve that.

Before I do that I'd like some more detailed information about what it takes to run a delegate.
- How much "stake" is needed to be eligible?
- How secure is that stake? Is there a way to secure my stake in the form of a mining multisig-wallet safely backed by my paper wallet(s) in a secure location or with the help of a cold-storage/signing solution?
- Any other details would be welcome like requirements in the shape of storage, cpu-power, memory, bandwidth etcetera.

 

 , also interested in this.

[HackFisher]

I'm not interested in running a delegate as a profession or company, but I'd be willing to help bootstrap the system and setup a delegate in the meantime, until the system has grown big enough.

I'd have to setup a cloudbased VPS for that though, because my home internet connection is rubbish and there's isn't much I can do to improve that.

Before I do that I'd like some more detailed information about what it takes to run a delegate.
- How much "stake" is needed to be eligible?
- How secure is that stake? Is there a way to secure my stake in the form of a mining multisig-wallet safely backed by my paper wallet(s) in a secure location or with the help of a cold-storage/signing solution?
- Any other details would be welcome like requirements in the shape of storage, cpu-power, memory, bandwidth etcetera.

 

[HackFisher]

For a delegate with high availability, you may need to handle DDOS attack to the DAC.
DDOS attack to the DAC could probably achieve through attacking the delegates.

delegates only sign blocks .. No need to expose you IP address .. as far as I can remember from Dans posts.
I may be wrong though

Yes, I said that because I'm wandering and having the presumption whether or not there are possibilities that delegates IP could be exposed by others since they are connected to network?

You certainly could run many nodes to avoid that, but which increase the cost of operation.

[JoeyD]

I'm not interested in running a delegate as a profession or company, but I'd be willing to help bootstrap the system and setup a delegate in the meantime, until the system has grown big enough.

I'd have to setup a cloudbased VPS for that though, because my home internet connection is rubbish and there's isn't much I can do to improve that.

Before I do that I'd like some more detailed information about what it takes to run a delegate.
- How much "stake" is needed to be eligible?
- How secure is that stake? Is there a way to secure my stake in the form of a mining multisig-wallet safely backed by my paper wallet(s) in a secure location or with the help of a cold-storage/signing solution?
- Any other details would be welcome like requirements in the shape of storage, cpu-power, memory, bandwidth etcetera.

[xeroc]

For a delegate with high availability, you may need to handle DDOS attack to the DAC.
DDOS attack to the DAC could probably achieve through attacking the delegates.

delegates only sign blocks .. No need to expose you IP address .. as far as I can remember from Dans posts.
I may be wrong though

[HackFisher]

If I get the chance to be a delegate from the beginning I would do it .. even if its not profitable at the beginning.

My Server is paid. I dont expect too much traffic from it and I also dont see why it should use too much processing power either

For a delegate with high availability, you may need to handle DDOS attack to the DAC.
DDOS attack to the DAC could probably achieve through attacking the delegates.

[xeroc]

If I get the chance to be a delegate from the beginning I would do it .. even if its not profitable at the beginning.

My Server is paid. I dont expect too much traffic from it and I also dont see why it should use too much processing power either

[HackFisher]

There might be a market between users and delegates if users can choose the amout of fee to pay. But maybe delegate could be voted down for the reason that they do not include low fee transactions.

[santaclause102]

Can't the pay for a delegate be dependent on the tx volume?

Pay is based upon trx volume and fee, but early on there may not be enough volume. 

Unless you want to subsidize delegates early on via inflation the pay may start out low.

What I meant was: When you have low tx volume you could give delegates a higher percentage of the total fees (10% by default). The relation of tx volume and % of the fees for delegates could be made variable to a degree.

[bytemaster]

Can't the pay for a delegate be dependent on the tx volume?

Pay is based upon trx volume and fee, but early on there may not be enough volume. 

Unless you want to subsidize delegates early on via inflation the pay may start out low.

[santaclause102]

Can't the pay for a delegate be dependent on the tx volume?

[bytemaster]

There is a very real chance that it will not be profitable to be a delegate in the early phases... however, establishing yourself as a proven delegate puts you in a good place once it becomes profitable.   Something we will have to figure out.

[luckybit]

U do not need 100%. 99.9% is good enough.  Also you can be down for 5 to 10 minutes between turns.

Remember you just need to be in the top 100 which is something anyone with a DO node could do.  Once compition heats up then you have to raise the bar.


Sent from my iPhone using Tapatalk

What if it's not profitable in the beginning? Not sure if it will make a profit if it depends on volume.

But there are probably ways to decrease costs.

[天籁]

    I will definitely support the Lab with all my XTS. Go,Go,Go!

[HackFisher]

OK, would like to provide a mail list for those who want to be in  to discuss related topics.

bitsuperlab@googlegroups.com

DPOS details whiter paper are already there:
http://bitshares.org/documentation/group__dpos.html

The first thing could be do researches or delegate software optimizations.

http://www.bitsuperlab.com/forum.html

[HackFisher]

OK, would like to provide a mail list for those who want to be in  to discuss related topics.

bitsuperlab@googlegroups.com

DPOS details whiter paper are already there:
http://bitshares.org/documentation/group__dpos.html

The first thing could be do researches or delegate software optimizations.

[testz]

  I'm in 

[Pocket Sand]

  This is something I'd definitely be more than interested in. Great idea 

[HackFisher]

I would like to donate a domain name as a start:

http://www.bitsuperlab.com

来自我的 GT-N7100 上的 Tapatalk

[bytemaster]

U do not need 100%. 99.9% is good enough.  Also you can be down for 5 to 10 minutes between turns.

Remember you just need to be in the top 100 which is something anyone with a DO node could do.  Once compition heats up then you have to raise the bar.


Sent from my iPhone using Tapatalk

[HackFisher]

The requirements exclude me personally. While I might have the bare minimum linux-skills to get the basics up and running, I have health problems that make it impossible for me to guarantee 100% up-time and responsiveness in case of trouble.

Should initial costs be to high, perhaps this could be crowdfunded/sourced or setup similarly to mining pools where a team maintains a couple of servers, like Amazon cloud-servers or similar, and members of  those pools share the costs and parts of the profits.

That is why we need to collaborate to form such a super lab!

I would like researches of this lab to be open, free to share, and high quality peer review, which could form an technical advantage as a team and be easy to fork, the other key points are depend on operation of mining servers

I'm not sure whether there are ways to securely share delegate secret key between multiple operators, may be we can find ways. Otherwise, we can also easily to fork to multiple delegate operators using similar strategy from DPOS Super Lab, but in this case, keeping it 100% up-time could be a problem.

[JoeyD]

The requirements exclude me personally. While I might have the bare minimum linux-skills to get the basics up and running, I have health problems that make it impossible for me to guarantee 100% up-time and responsiveness in case of trouble.

Should initial costs be to high, perhaps this could be crowdfunded/sourced or setup similarly to mining pools where a team maintains a couple of servers, like Amazon cloud-servers or similar, and members of  those pools share the costs and parts of the profits.

[boombastic]

would like to help


Sent from my iPhone using Tapatalk

[HackFisher]

This is exactly the benefit of DPOS... 100 highly reliable nodes makes the 'nodeshare' concept that BTC is facing unnecessary.

Big money to be made here by providing this service to the network.

Yes, that's it. And BM has given a comprehensive summary of the requirements as following:

Becoming a delegate requires:

1) Ability to run provided software on a low-latency connection 
2) Ability to setup a secure linux environment
3) Ability to get 1% of the vote...
      a) unique geographic location
      b) trusted member of the community
      c) 1000 BIPS  (a couple of PTS or AGS) registration fee... (keep out spammers)

Remaining a delegate requires:
      a) low latency, 99.9% uptime
      b) non-discrimination of valid transactions
      c) better reputation than all but the top 99

How to earn the best reputation:
      a) Provide the most additional services for your pay.
            - seed node supporting thousands of connections
            - block explorers
            - free web wallets
            - high speed blockchain downloads
      b) Provide the most proof as to your transparency
     

You get the idea... who ever is willing to provide the most service for their pay will likely win over someone attempting to provide the least service. 
 
     

As I understand, to prevent from DDOS for 99.9% uptime, the server might need to be distributed around the world, have 24 hour clock team to operate, high quality mining software etc.

Brainstorm might be needed first.

[bytemaster]

This is exactly the benefit of DPOS... 100 highly reliable nodes makes the 'nodeshare' concept that BTC is facing unnecessary.

Big money to be made here by providing this service to the network.

[xeroc]

.. im in

[toast]

would join this effort

Sent from my SCH-I535 using Tapatalk

[HackFisher]

After reading the DPOS whiter paper, I have an idea to form a super lab to join the competition of DPOS mining, because providing good service as a stable DPOS delegate is a challeging work, and could be very interesting and profitable.

Two goals:
1. Act as one of the 100 delegate mining nodes, and keep stable with high reputation, 100% online.
2. Testing DPOS network, reseach on all kinds of attack and cheating possibility. Find the flaw of other 99 mining nodes.

Any ideas, or anybody intent to collaborate?

来自我的 GT-N7100 上的 Tapatalk