v0.10.16 - Onyx v2

Another find.

When any client start it ask at least 3 other nodes for full MN list https://github.com/darkcoin/darkcoin/blob/master/src/darksend.cpp#L1978
and get them by dsee messages https://github.com/darkcoin/darkcoin/blob/master/src/masternode.cpp#L260
While first response will pass this loop https://github.com/darkcoin/darkcoin/blob/master/src/masternode.cpp#L90
without entering there and will keep lastTimeSeen equal to received lastUpdated https://github.com/darkcoin/darkcoin/blob/master/src/masternode.cpp#L141
next responses from other 2+ nodes will go inside loop and override lastTimeSeen with current time https://github.com/darkcoin/darkcoin/blob/master/src/masternode.cpp#L95
giving every masternode another 70 minutes of "life" on this particular client. So when someone else will start his client and connect to this client to get MN list he will refresh time for every masternode in his own list again.
I think MNs could "drift" quite a while by this feature :)
Solution:
modify this line https://github.com/darkcoin/darkcoin/blob/master/src/masternode.cpp#L95
to
Code: 
mn.UpdateLastSeen(lastUpdated);
and/or
move it inside next "if" https://github.com/darkcoin/darkcoin/blob/master/src/masternode.cpp#L97

EDIT: http://jira.darkcoin.qa/browse/DRK-126
 
Last edited by a moderator:
camosoul in another thread
camosoul said:
I've got 4 that I absolutely cannot get to show up in masternode list. They worked fine initially, changed nothing, refuse to show up. It would be nice if they weren't just ignored by regexp, but were still shown, as unlisted, in purple...

OT, but found reason; local client start-many is falsely claiming:

Code: 
No sutable vin found

This statement is false. All addresses were and still are locked. They all have exactly the same 1000 DRK they started with. Nothing changed. start-many is failing to recognize that which it previously had no problem recognizing.

Restart local client.

masternode start-many over and over again, it pick up a missing node randomly on one pass or another. After re-running it about 30 times (giving time for propagation, checking masternode list from the nodes themselves), all are back in the list.

Sumthin's goofy...
Click to expand...
Switching here....

The reason is start-many is quite..... simple right now :)
It reads file line by line https://github.com/darkcoin/darkcoin/blob/master/src/rpcdarksend.cpp#L211
and tries to run another masternode.
But it skips vins that are already in masternode list https://github.com/darkcoin/darkcoin/blob/master/src/activemasternode.cpp#L227
So if you have some of them already running then line count in your file will be more then count of available vins and you'll get this message.

As Evan said it's undocumented and experimental so you have it AS IS :wink:

But having "stop-many" command would help to solve such issues I guess....

EDIT: http://jira.darkcoin.qa/browse/DRK-127
 
UdjinM6 said:
camosoul in another thread

Switching here....

The reason is start-many is quite..... simple right now :)
It reads file line by line https://github.com/darkcoin/darkcoin/blob/master/src/rpcdarksend.cpp#L211
and tries to run another masternode.
But it skips vins that are already in masternode list https://github.com/darkcoin/darkcoin/blob/master/src/activemasternode.cpp#L227
So if you have some of them already running then line count in your file will be more then count of available vins and you'll get this message.

As Evan said it's undocumented and experimental so you have it AS IS :wink:

But having "stop-many" command would help to solve such issues I guess....
Click to expand...

I look forward to it being made more robust and to stop-many. I realize these are not front-burner projects right now and require a little sniggling. Just providing my sniggling so someone else can use it if they need it.
 
djcrypto said:
Show me how?
Click to expand...
Set up sessions in Putty for your ip addresses of your masternodes. Then load those sessions into superputty and organize the windows so they are all built off the first instance. For instance, session 1 would have the ability to add session 2 above, below, left, or right of that session. Continue building off these sessions, then use the above command line to send commands to all the masternodes at once.
 
how do we know when the latest update is a protocol update and requires re-starting the local wallet? Is it when the third # in the build changes, like 0.11.15 to 0.11.16?

Thanks!
 
oblox said:
Set up sessions in Putty for your ip addresses of your masternodes. Then load those sessions into superputty and organize the windows so they are all built off the first instance. For instance, session 1 would have the ability to add session 2 above, below, left, or right of that session. Continue building off these sessions, then use the above command line to send commands to all the masternodes at once.
Click to expand...

this works great, thanks for the tip of using SuperPutty!
 
A
Wh1teKn1ght said:
how do we know when the latest update is a protocol update and requires re-starting the local wallet? Is it when the third # in the build changes, like 0.11.15 to 0.11.16?

Thanks!
Click to expand...

Actually, this is an excellent question. I don't know how the protocol number would help (mine says "protocolversion" : 70046 - how would that tell you?) So for me, I try to update the remote server (the one with the ip address, for me -run by amazon) and I give it half a minute or so. Usually, it will say masternode successfully started or some such right away. Sometimes it takes a bit of time, but not long. If it doesn't automatically detect the masternode starting, then I start up the local masternode. It really all depends on what was changed though. I don't think there is a clear answer.
 
TanteStefana said:
I don't know how the protocol number would help (mine says "protocolversion" : 70046 - how would that tell you?)
Click to expand...

darkcoind getinfo:

  1. "version" : 101612: this is the software version you use
  2. "protocolversion" : 70046: this is the version of the protocol for communication
As long as protocolversion stays the same you only need to restart your hot Masternode when you install a new version.
If it changes you have to start the Masternode via the cold wallet with 'darkcoind masternode start "<your_passphrase>" ' .
 
Last edited by a moderator:
ouch: collaterals! (top wallet balance was static, bottom wallet had deposits occur across earlier ds rounds )
six fees occured on bottom wallet
(no internet outages, no stop/starting darksend)
both win-qt v0.10.16.12-gc588770-beta

on0Pjel.png
 
I think start-many screwed things up for me. Now I'm trying to get my mns working one at a time, but I don't know how to get them to show up on https://drk.mn . I stop both remote and local, remove peers.dat on both, start remote, and masternode start <passphrase> on local works fine, and ./darkcoind masternode list | grep <ip address> does show it as active "<ip address>:9999" : 1, but it never appears on https://drk.mn :sad:

./darkcoind masternode count the number keeps changing. ??

I keep doing the same thing over and over, wasting so much time.


edit: Most of my mns that currently do appear on https://drk.mn show the wrong public key for the ip address, probably because start-many screwed things up?
 
Last edited by a moderator:
moocowmoo said:
ouch: collaterals! (top wallet balance was static, bottom wallet had deposits occur across earlier ds rounds )
six fees occured on bottom wallet
(no internet outages, no stop/starting darksend)
both win-qt v0.10.16.12-gc588770-beta

on0Pjel.png
Click to expand...

I just finished !
Good timing !
Tx moocowmoo
 
moocowmoo said:
ouch: collaterals! (top wallet balance was static, bottom wallet had deposits occur across earlier ds rounds )
six fees occured on bottom wallet
(no internet outages, no stop/starting darksend)
both win-qt v0.10.16.12-gc588770-beta

on0Pjel.png
Click to expand...

There is no "collateral" anymore. You will be charged 0.1 "randomly" while you are using Darksend. This is to prevent DOS and other Sybil like attacks from occurring for free.

The charge should be rare according to Evan in the first post, so it looks like you had a run of bad luck.

Of course it could be a bug...
 
You must log in or register to reply here.
Back
Top