Bible Pay

Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Messages - Rob Andrews

Pages: 1 ... 231 232 233 234 235 236 237 [238] 239 240 241 242 243 244 245 ... 263
3556
Looks like we have 15 CPIDs crunching in Test!  Awesome.

Lets keep going, lets shoot for 50!  This way we can huddle together in a few weeks and make a determination call on PODC- to go with it in prod or not go with it. 


3557
If you have ever had any blocks or chainstate prior to v1.0.8.8, you probably got banned.
Please try:
rm chainstate -r
rm blocks -r

Then resync.  Im up now so you should be able to resync now.
Btw:  Please do this in the \testnet3 folder, not in the prod folder.


3558
myself and the 5 peers i'm connected to are all running 1.0.8.9, but after my most recent restart there's no blocksource available and i'm only up to block 6016.

If you have ever had any blocks or chainstate prior to v1.0.8.8, you probably got banned.
Please try:
rm chainstate -r
rm blocks -r

Then resync.  Im up now so you should be able to resync now.


3559
got my researcher/wallet running yesterday in testnet, confirmed payments this morning with getboincinfo and transaction history.  one thing i've had happen a couple times is that the wallet has banned all the peers i'd been connected to and i have to restart the wallet and unban everyone to reconnect to the network.

any thoughts?

I think I know whats going on - Im the main seed node (node.biblepay.org) in testnet and Im down, Im currently adding a new rule in the code.

There are many old testers on old versions.  Did you sync to block 6076?  Try clicking each peer and see if the are on 1089 in the gui version?
If so your good for now, but Ill probably be down for one more hour.  Then ill give you more live insight.


3560
Sounds good. are you expecting some to not be voted in?

In Prod:  Rarely.  I think with 130 sancs, as long as the quorum percent is correct (10%, thats 13 votes) then I think they will be voted in. 
 If Rosetta or boinc go down in prod, I have some advanced rules set up.  If the file is either corrupted, or empty, or unfilterable, or contains no magnitude or less than X researchers in prod (thats 5 right now in the code), the file hash is set to 0x0, and the sancs will not vote on 0x0.  So that is the most likely scenario in prod if we miss a superblock, maybe it will happen once a year.

In Test:  We started out with a 65% hit ratio (I added the feature and just ran it) and thats primarily because we didnt tell the sancs to run in -qt mode.   We can monitor it and see if we rise in test.




3561
I think tomorrow I will add:

Total payments to CPID over 7 day period, and a list of superblocks with payments comma separated over 7 days, those two fields to exec getboincinfo.

Then it should pretty much tell us the accuracy level we are hitting.

I could also add a ratio column : superblock hit vs not hit. (IE Hit Ratio : 75%).  That just tells us how many failed to be voted in.




3562
Rob, OK I understand now. Thanks for showing me how to check on my own. Good work on the integration by the way.

Thanks dude.

3563
Yeah, I would like to get to the bottom of my missing payment, not sure what went wrong. It could be that I read wrong but...

Maybe we misunderstood each other - I am saying you definitely didnt see a magnitude earlier, because you are not (definitely not) in the SAN/filtered file.

If you want to see the file, please cd /SAN and cat filtered.

It only grows in size, it does not shrink.

Therefore you were never with magnitude, and therefore would have no payments.

Your CPID association was < 24 hours ago, you are last in the list of dccs:  exec list dccs

To help cut the confusion down, I just added a feature to show the Warning I mentioned in the previous post.




3564
It's at 247 now. I think we need to have better status update or we will have lots of the same questions to answer regarding RAC and payment. When I said I saw a RAC number I meant I saw it on the front page.


02:33:37

exec getboincinfo


02:33:38

{
  "Command": "getboincinfo",
  "CPID": "e94c1704c75f731f8bfde303f08408ee",
  "Address": "yiqLtGKtrDykVqrFREoZ8B5mhRE66s4cw8",
  "CPIDS": "e94c1704c75f731f8bfde303f08408ee;",
  "NextSuperblockHeight": 5643,
  "e94c1704c75f731f8bfde303f08408ee_RAC": 247.62,
  "LastSuperblockHeight": 5247,
  "Payments": 0,
  "Budget": 1386000,
  "Magnitude": 0
}

I think what we need is a sentence saying Please wait 24 hours for magnitude to populate.... I think I can query when a person associated their CPID, and if its < 24 hours, maybe place a warning in the getboincinfo : Warning CPID associated 14 hours ago, please wait 24 hours for magnitude to populate.  If the magnitude is 0 on the front page, thats normal for the first 24 hours.


3565
CPID: e94c1704c75f731f8bfde303f08408ee

Before I took the node down about 3 hours ago I saw a number on the front page but now it's 0.
I think you saw your CPID appear on the front page but not magnitude, since you are still not in the san\filtered file yet (that will always only get bigger).  Yeah your not in yet, so you should have 0 payments.

Also, 'exec getboincinfo' shows your up to date RAC and should show 0 mag.  Do u have RAC yet?


Maybe we should move to once per day superblocks in testnet so we dont have to check as many.  We need some type of report you can run that shows every cpid and payment in the last superblock.




3566
Rob, when my mag actually had a number earlier today, I didn't see a payment for it.  How do I check?
If your mag had a number earlier it should still have a number, why no number now, did you move wallets?

The easiest way is to just scroll through your transaction history, look for anying over 20k in the last 24hours?

Since we have 7 superblocks a day, you'd really have to figure the spacing math (exec getboincinfo, see the difference between the 2) then showblock each prior block for 24 hours to check manually, but from what I remember, you didnt have a mag all day, this morning you were not in the list, what is your cpid?


3567
Mine says:
AcceptBlock::ExecuteDistributedComputingSanctuaryQuorumProcess PENDING_SUPERBLOCK .
2018-02-10 02:05:00 UpdateTip: new best=6ee2dc9bc66a6310009d192ea9c6f67a4bd29e8efa17f92e2046374f8311281e  height=5571  log2_work=39.043051  tx=7504  date=2018-02-10 02:04:59 progress=0.031810  cache=0.1MiB(594tx)
2018-02-10 02:05:00 AddToWallet 5b3f2d75a103ead2d58a5ef32a41a41ff00e48d10b3f9726e0a92b915c3f21d9  new
2018-02-10 02:05:00 CMasternode::UpdateLastPaidBlock -- searching for block with payment to 8b2affbe3d46b48ec87e6e6115f06e511a9673697f647b8ef1aa25172e14175b-1 -- found new 5571.000000
2018-02-10 02:05:00 ProcessNewBlock : ACCEPTED

Sweet,  so really things seem to be working properly.

Tomorrow I will move the distinct CPID rule to the blockcheck for testnet and have a release (Ill wait til something else can be merged with it before releasing another testnet release).

Does anyone have any security questions, or any questions about anything DC related?

Maybe I should add a feature to show total payments over 7 days, then we can do some math also.


3568
Ok, Im back and I see 2 of you did vote on the contract while I was down, which is great, in last superblock.
Why dont we try checking this next contract out without forcing it?

Mike are you online?
So I kind of answered my own question.  What I meant by force, was if users keep typing exec testvote, they are forcing a vote for their contract hash (which the code will do anyway, once per block).
So checking to see what kind of support we have for the  next superblock (without typing exec testvote),  you can type
cat debug.log | grep ExecuteDistributed -A5
And you will see what your node thinks its doing as blocks click by.  In my case it says "PENDING SUPERBLOCK", meaning two Sancs already voted on block #5643, telling me there will be a superblock.

Each Sancs vote only counts once per filehash.



3569
Have 2 masternodes running on 1.0.8.9 now.

Ok, Im back and I see 2 of you did vote on the contract while I was down, which is great, in last superblock.
Why dont we try checking this next contract out without forcing it?

Mike are you online?

3570
Have 2 masternodes running on 1.0.8.9 now.
I accidentally spent my hot masternode funds...recreating now.

Pages: 1 ... 231 232 233 234 235 236 237 [238] 239 240 241 242 243 244 245 ... 263