Dry Run 8: Need for Speed

[Fox]

FOX is testing enabling the same delegate on two nodes to attempt to fork/attack the network.

Going to pull latest code and build on three nodes on different continents in an attempt to fork based on geo-separation.

Good news: I was unable to introduce a sustained fork by running four (4) clients in three (3) separate regions. Therefore I have returned to normal single delegate node for block production.

In Dry Run 9 I'd like to modify the test by increasing the number of delegates producing blocks from multiple nodes. The thought being that consecutive block production from a given series of nodes could produce a blockchain that differs from a separate set of nodes where peering does not support efficient transaction proposition. Given, the delegates producing from multiple nodes must do do consecutively which is by design random per round, so it would take a large set of multi-node delegates for even the possibility of the 'test' going to plan.

Is this a potential attack vector, malicious or otherwise?

Thoughts?

[bitcoinerS]

I tried running dev_tests, but it seems to get stuck here. Anyone else having this problem?

cli.cpp:1476                  wait_till_cli_shutdo ] waiting on server to quit

Code: [Select]

* Short Order

A: blockchain_market_order_book GLD XTS
blockchain_market_order_book GLD XTS
1491516ms th_a       cli.cpp:178                   execute_command_line ] line: blockchain_market_order_book GLD XTS
1491516ms th_a       chain_database.cpp:2351       get_market_asks      ] end of db
1491516ms th_a       asset.cpp:239                 operator*            ] r.amount = 4000000
1491517ms th_a       asset.cpp:240                 operator*            ] {"amount":100000,"asset_id":2} * {"ratio":"0.025","quote_asset_id":2,"base_asset_id":0} => {"amount":4000000,"asset_id":0}
1491517ms th_a       asset.cpp:214                 operator*            ] {"amount":34500000,"asset_id":0} * {"ratio":"0.0656","quote_asset_id":2,"base_asset_id":0} => {"amount":2263200,"asset_id":2}
1491517ms th_a       asset.cpp:214                 operator*            ] {"amount":271493,"asset_id":0} * {"ratio":"0.0221","quote_asset_id":2,"base_asset_id":0} => {"amount":5999,"asset_id":2}
1491517ms th_a       asset.cpp:214                 operator*            ] {"amount":21300000,"asset_id":0} * {"ratio":"0.0767","quote_asset_id":2,"base_asset_id":0} => {"amount":1633710,"asset_id":2}
1491517ms th_a       asset.cpp:214                 operator*            ] {"amount":419580,"asset_id":0} * {"ratio":"0.0143","quote_asset_id":2,"base_asset_id":0} => {"amount":5999,"asset_id":2}
1491517ms th_a       asset.cpp:214                 operator*            ] {"amount":12000000,"asset_id":0} * {"ratio":"0.08","quote_asset_id":2,"base_asset_id":0} => {"amount":960000,"asset_id":2}
                               BIDS                                |                                ASKS                               
TOTAL                     QUANTITY                           PRICE | PRICE                           QUANTITY                     TOTAL
---------------------------------------------------------------------------------------------------------------------------------------
100.000 GLD               40.00000 XTS               2.5 GLD / XTS | 6.56 GLD / XTS             345.00000 XTS             2,263.200 GLD
5.999 GLD                 2.71493 XTS               2.21 GLD / XTS*| 7.67 GLD / XTS             213.00000 XTS             1,633.710 GLD
5.999 GLD                 4.19580 XTS               1.43 GLD / XTS*| 8. GLD / XTS               120.00000 XTS               960.000 GLD

* Short Order

1491520ms th_a       cli.cpp:1476                  wait_till_cli_shutdo ] waiting on server to quit



[crazybit]

seems the Dry Run 8 latest client consumes too much memory, 2G memory would be used up soon after i bring up the client in ubuntu14. and it worked fine without memory issue in the previous rounds of dry run test, do not know why it consumes so much memory now, is there any memory leak we do not know yet?and anybody encountered the similar issue?

[CalabiYau]

The seed nodes dropped almost all connections and we have those logs as well.

Good to know, thx

[bytemaster]

The seed nodes dropped almost all connections and we have those logs as well. 

[CalabiYau]

All, I have re-enabled network multi-threading by default in light of recent fixes. Seems to be running fine for me so far (just some slowness initially connecting). Please pull and build the latest code and see if it gives you any problems.

pm sent to vikram with some p2plog data.

Code: [Select]

  "bitshares_toolkit_revision": "ca799937c4f2bc09f5b3042e916374f80c9d0872",
  "bitshares_toolkit_revision_age": "11 hours ago",
  "fc_revision": "de3e2610a5033a766931b5873a26ac8b122cdbcb",
  "fc_revision_age": "15 hours ago",
  "compile_date": "compiled on Jul 13 2014 at 08:19:55"

Still lot of unfulfilled wishes 

Code: [Select]

Still want to connect to more nodes, but I don't have any good candidates.  Trying again in 15 seconds


Code: [Select]

--- there are now 4 active connections to the p2p network
--- in sync with p2p network
--- there are now 0 active connections to the p2p network
--- there are now 1 active connections to the p2p network
--- there are now 2 active connections to the p2p network
--- there are now 3 active connections to the p2p network
--- in sync with p2p network
--- there are now 2 active connections to the p2p network
--- there are now 3 active connections to the p2p network
--- in sync with p2p network
--- there are now 0 active connections to the p2p network
--- there are now 1 active connections to the p2p network
--- there are now 2 active connections to the p2p network
--- there are now 3 active connections to the p2p network
--- in sync with p2p network
--- there are now 0 active connections to the p2p network
--- there are now 1 active connections to the p2p network
--- there are now 2 active connections to the p2p network
--- in sync with p2p network
--- there are now 3 active connections to the p2p network
--- in sync with p2p network


[liondani]

maybe only the last logs?

[sumantso]

Lost all connections again

Save your p2p logs and post them somewhere.  I'll have Eric look into this.


Sent from my iPhone using Tapatalk

Just checked the folder - its 1 GB! I will try to upload it somewhere.

[bytemaster]

Lost all connections again

Save your p2p logs and post them somewhere.  I'll have Eric look into this.


Sent from my iPhone using Tapatalk

[sumantso]

Lost all connections again

[BTSdac]

what is configuration of computer to run a delegates?
I have a delegate ,can not produce block while clients are runing in VPS , but can 100% produce block while clients are ruing in my personal computer . but I cannot make my comuter are been runing 24*7,  so what is the Min configuration of computer if I want to run a delegate in VPS .

[bytemaster]

In the Dry Run 9 branch I have created the first market-issued BitUSD... next dry run will give you all a chance to try it out.

I still a few things to implement:

1) Automatic Covering
2) RPC calls for managing your short positions
3) Proper handling of market fees.
4) Minimum Market Depth Enforcement
5) Initial Market Depth Enforcement
6) Maximum Price Change per block.

None of these items should take very long so we should see some significant progress this week. 

[Fox]

FOX is testing enabling the same delegate on two nodes to attempt to fork/attack the network.

Going to pull latest code and build on three nodes on different continents in an attempt to fork based on geo-separation.

[vikram]

FOX is testing enabling the same delegate on two nodes to attempt to fork/attack the network.

Roger that.

All, I have re-enabled network multi-threading by default in light of recent fixes. Seems to be running fine for me so far (just some slowness initially connecting). Please pull and build the latest code and see if it gives you any problems.

[bytemaster]

FOX is testing enabling the same delegate on two nodes to attempt to fork/attack the network.