KanoPool since 2014 🐈 - PPLNS and Solo 0.5% fee - Worldwide - 2435 blocks

[ZACHM]

Old is when you did this on your parents analoge landline 2 wire copper phone connection:

https://en.wikipedia.org/wiki/Acoustic_coupler

to e.g. use BBS's via X.25 /x.28 providing similar services as e.g. this forum does... just whithout mouse clickee and graphics... 300 bps max...

(there where Nets bigger and more interesting than Internet at that time )

 

really getting old...

 

Woo hoo, I'm not old, but close. My first "PC" was a 486 computer that had the sweet "turbo" button, 512mb hard drive. Installed a 14.4k modem and voila the WWW.  

Before that I had vic20, commodore 128, etc.

[1714814506]

1714814506

[1714814506]

1714814506

[1714814506]

1714814506

[1714814506]

1714814506

[-ck]

did we loose one

This block 388788 is showing up as an F2Pool block on the Blockchain...  

Yeah just doing an analysis of what happened.
They mined 2 blocks quickly and thus orphaned ours.
More info later once I work out the timings.

I discussed this case in a different post months ago, but when your pool gets a block then your pool is the first one to start on the next block and if you get lucky and that next block is ultra low diff then you will likely get it before the rest of the miners start mining it, guess it a 100-200ms advantage.

Turns out it was nothing like that. The share from the miner was just a very stale share. Had it come 2 minutes earlier it would have been a block but it was over a minute late. The miner has some very old hardware in his mix so it's possible the device in question was very bad at working on fresh work.

[wikkidtt]

Old is when you did this on your parents analoge landline 2 wire copper phone connection:

https://en.wikipedia.org/wiki/Acoustic_coupler

to e.g. use BBS's via X.25 /x.28 providing similar services as e.g. this forum does... just whithout mouse clickee and graphics... 300 bps max...

(there where Nets bigger and more interesting than Internet at that time )

 

really getting old...

 

Woo hoo, I'm not old, but close. My first "PC" was a 486 computer that had the sweet "turbo" button, 512mb hard drive. Installed a 14.4k modem and voila the WWW.  

Before that I had vic20, commodore 128, etc.

Mine was a chisel and stone beat that!

[VirosaGITS]

So, we just got a share that would of found us a block, but it was stale. The tooltip says "Share was submitted 1 minute too late." Does this mean the miner submitted too late to the pool late, or did we simply legit lost the propagation race?

[d57heinz]

Old is when you did this on your parents analoge landline 2 wire copper phone connection:

https://en.wikipedia.org/wiki/Acoustic_coupler

to e.g. use BBS's via X.25 /x.28 providing similar services as e.g. this forum does... just whithout mouse clickee and graphics... 300 bps max...

(there where Nets bigger and more interesting than Internet at that time )

 

really getting old...

 

Woo hoo, I'm not old, but close. My first "PC" was a 486 computer that had the sweet "turbo" button, 512mb hard drive. Installed a 14.4k modem and voila the WWW.  

Before that I had vic20, commodore 128, etc.

Mine was a chisel and stone beat that!

I remember growing up my dad chatting on packet radio.  Was a 9600 baud modem connected to ham radio   He ran the local 146.820 repeater.  You could chat with an apple 2 e and this modem to anyone within range of the repeater.  That was back in the early 80s


Best regards
D57heinz

[-ck]

So, we just got a share that would of found us a block, but it was stale. The tooltip says "Share was submitted 1 minute too late." Does this mean the miner submitted too late to the pool late, or did we simply legit lost the propagation race?

The miner should never have been working on the stale work in the first place. The pool had informed all miners of the block change and the need to change work 69 seconds earlier. Older mining hardware/software is probably the reason - bear in mind this doesn't mean this miner is getting paid for stale work as his rejects are higher and they're not paid for.

[kano]

So, we just got a share that would of found us a block, but it was stale. The tooltip says "Share was submitted 1 minute too late." Does this mean the miner submitted too late to the pool late, or did we simply legit lost the propagation race?

It wasn't an orphan after all, it was a stale share.

He submitted it 69 seconds late to kano.is
ckpool has (as yet before never tested) code to try submit stale block shares anyway ... just in case.

kano.is put out our 388788 block change to F2Pool's block on time:
[2015-12-17 11:38:04.925] Block hash changed to 00000000000000000497f638fc378166d2d14742d4b77187209fe5f32a57afe3

poolbench also shows they saw it 573ms later:
stratum.kano.is:3333 2015-12-16 19:38:05.498000

But the share came in 69 seconds later:
[2015-12-17 11:39:13.492] Possible block solve diff 182123374943.539276 !
[2015-12-17 11:39:13.604] SUBMIT BLOCK RETURNED: inconclusive
[2015-12-17 11:39:13.604] Submitted, but rejected block 388788

So yeah there was no chance in hell of it ever being a block

Edit: oh also CKDB's report on it (including my manual interventions 'Orphan' and 'Unworthy')

Code:

[2015-12-17 11:39:13.515+11] blocks_add(): BLOCK! Status: New, Block: 388788/...00006097e9a3b4d0 Diff 182G UTC:2015-12-17 00:39:13.492270+00
[2015-12-17 11:39:13.515+11] BLOCK! STALE Diff 230.2% (182123374944/79102380900.2) jbravo.rpi0166b Pool 92112911921.0 92.1G 116.45%
[2015-12-17 11:48:05.275+11] make_a_shift_mark() mark 6228334527803037660/e/x/Shift fin: 5ppdo charlotte/Block 388788 fin/
[2015-12-17 11:48:13.327+11] SS_to_MS() Processing: workmarkers 20140918511339//End 6228334527803037660/Stt 6228334527803037601/Shift fin: 5ppa6 belldandy++ to Shift fin: 5ppdo charlotte/x
[2015-12-17 11:48:15.960+11] SS_to_MS() Processed: 3684 ms 218670 ss 1817804 shares 2391717373 diff for workmarkers 20140918511339//End 6228334527803037660/Stt 6228334527803037601/Shift fin: 5ppa6 belldandy++ to Shift fin: 5ppdo charlotte/x
[2015-12-17 11:56:20.737+11] cmd_query() 1.block height=388788
[2015-12-17 11:56:57.677+11] cmd_query() 1.block height=388787
[2015-12-17 11:58:29.852+11] blocks_add(): Status: Orphan, Block: 388788/...00006097e9a3b4d0 Diff 182G
[2015-12-17 11:58:37.627+11] summarise_blocks() block 388788, stats confirmed 91908114882/297315718/77826749/276997/78216
[2015-12-17 11:58:38.628+11] process_pplns(): can't process block 388788/6228334527803037660/jbravo.rpi0166b/2559600823 status: O/Orphan
[2015-12-17 12:06:34.142+11] blocks_add(): Status: Unworthy, Block: 388788/...00006097e9a3b4d0 Diff 182G

[VirosaGITS]

So, we just got a share that would of found us a block, but it was stale. The tooltip says "Share was submitted 1 minute too late." Does this mean the miner submitted too late to the pool late, or did we simply legit lost the propagation race?

The miner should never have been working on the stale work in the first place. The pool had informed all miners of the block change and the need to change work 69 seconds earlier. Older mining hardware/software is probably the reason - bear in mind this doesn't mean this miner is getting paid for stale work as his rejects are higher and they're not paid for.

Thank you very much for the explanation. I get it now. It is unfortunate that this may happen because a miner's software did not react properly to the job change notification.

Being one of our (new(?)) biggest miners here at Kano's C/K Pool, i hope jbravo will fix anything there is to fix, if there is something to fix and if it was simply just not a fluke.

@Kano i see your reply just now. Thank you for the in depth log. Not changing job 1 min after a job change seem like the miner would be wasting a lot of shares.

[not.you]

Look at the Workers->Shift Graph
The light red shading is the 5Nd+ of the last reward

So now that you took the time to make that nice shading you need to add to your "to do" list to make it unshade for orphans and stale blocks  

Also now that I see it, eventually it will all be shaded except for the leading edge so maybe each block should get a different color?

[wcg]

What sucks is all the time and power involved and we all get nothing

[Nuttycoins]

Something smells of stale fish to me....

[ZACHM]

What sucks is all the time and power involved and we all get nothing

None of the time or power is lost, it is no different than if it had been a low diff share. Everything just keeps going towards the next block.

[kano]

Look at the Workers->Shift Graph
The light red shading is the 5Nd+ of the last reward

So now that you took the time to make that nice shading you need to add to your "to do" list to make it unshade for orphans and stale blocks  

Also now that I see it, eventually it will all be shaded except for the leading edge so maybe each block should get a different color?

Yeah it doesn't actually have the marks, it calculates (guesses?) them from the shift data.
So in this case it saw a 'Block' marker and decided that's where it starts on the right.
It doesn't know if the block was actually a block or not.
That's what I meant when I said:

...
Yeah I'll put a label on it and also put a 'wild guestimate' of the next range on the right somehow in the not too distant future.
- maybe just use the same day range size as the last payout and mark that from the right edge back to the left
That will be wrong if the hash rate changes, but probably better than nothing.
But that will require a ckdb change to actually provide the correct last payout details - I currently find it in the shift information, if it's there, which isn't guaranteed.

[generalt]

Thanks for the analysis.  It's nice to have pool ops that go through the details to its miners rather than just be left in the dark like most other pools.

[citronick]

So, we just got a share that would of found us a block, but it was stale. The tooltip says "Share was submitted 1 minute too late." Does this mean the miner submitted too late to the pool late, or did we simply legit lost the propagation race?

The miner should never have been working on the stale work in the first place. The pool had informed all miners of the block change and the need to change work 69 seconds earlier. Older mining hardware/software is probably the reason - bear in mind this doesn't mean this miner is getting paid for stale work as his rejects are higher and they're not paid for.

Thank you very much for the explanation. I get it now. It is unfortunate that this may happen because a miner's software did not react properly to the job change notification.

Being one of our (new(?)) biggest miners here at Kano's C/K Pool, i hope jbravo will fix anything there is to fix, if there is something to fix and if it was simply just not a fluke.

@Kano i see your reply just now. Thank you for the in depth log. Not changing job 1 min after a job change seem like the miner would be wasting a lot of shares.

in other pools, some have separate ASICs only pool to facilitate new fast miners from the non-ASICs one - would this help the above?

[-ck]

in other pools, some have separate ASICs only pool to facilitate new fast miners from the non-ASICs one - would this help the above?

There's no such thing as non-asic miners for bitcoin.

[clgrissom3]

Thanks for the analysis.  It's nice to have pool ops that go through the details to its miners rather than just be left in the dark like most other pools.

Agreed!  It's nice to be included and to be a part of something like this.  There is no communication out in the other pools...when something looks out of place or is going wrong you get nothing but crickets…chirp…chirp.

[bctmke]

Saw that we had hit a block and hopped on the forums quick.  Realized very quickly that it wasn't (hadn't actually looked at the pool stats yet).  It is great that -ck and kano will pop on and let us know whats up.  Here's to the next one!

[wcg]

Is the factor cg miner the best for mining on this pool do you recommend an update or somthing that will perform better
I have s7 and s5 miners ...thanks

[wolfen]

Is the factor cg miner the best for mining on this pool do you recommend an update or somthing that will perform better
I have s7 and s5 miners ...thanks

Pretty sure Kano will handle your hardware with stock firmware.
S7's working top notch.
Pool won't slam your rigs around.