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

[VRobb]

A block at less than 100% diff!  Finally, let's get some green to go with our red for more holiday spirit!!

[1714825791]

1714825791

[1714825791]

1714825791

[1714825791]

1714825791

[1714825791]

1714825791

[philipma1957]

yeah I could use some block action,

So the last two blocks were 103 and 90  that is okay

[not.you]

is s9v2 the autotune firmware?? I am wanting to upgrade mine from v1

Well, in case it wasn't obvious ...
I have no idea about what firmware runs in what S9
So I didn't answer before, but maybe I should bump this for someone else to answer?

To be explicit about it (I think it was sort of answered but not directly), the BMminer v 1.0 firmware is the non-autotune firmware.  The autotune firmware is BMminer v 2.0.

The autotune firmware takes away under\overclocking (as mentioned) but it also takes away manual fan speed control which is a bit lame.  This of course only applies to upgrading the firmware on the earlier models that came with the non-autotune firmware, because if you bought a later model that came with the autotune firmware you never had control of those things anyway.  You cannot downgrade a newer model to a non-autotune firmware but you can upgrade an older model to the autotune firmware, but it will not autotune hashboards that are not designed for autotune.  So basically if you upgrade an older model it will just run at whatever frequency you tell it is stock.  Blockchainmines mentioned this already but when you install the autotune firmware on an older model it asks you to set the stock freq and then it runs at that.  If you mix and match hashboards (like if one breaks and you get a replacement) then the new style hashboards will autotune and old ones will not within the same miner (meaning that the firmware will autotune the boards that are capable of it and run the others at whatever it was told stock freq is).

The earlier versions of the auto-tune firmware had fan speed issues that made the speed go up and down like a yo-yo, the newest versions are more stable in terms of fan speed.  It is still a gamble though to upgrade, I have seen a few accounts were people had serious problems after trying to put the autotune firmware on an older S9.

[ComputerGenie]

I don't know about auto-tune but it sucks!  You have one shot to pick your frequency and then it is locked in and that tab disappears in the settings.  The only way to try different frequency to find the optimal setting is to re-flash the entire device and choose another frequency and subsequently get re-locked in.  I myself am a default kinda guy.  I don't overclock, or mess with the settings along as it hashes like it supposed to.  However in some cases I have found on the S9's or all antminers I have used since the S2 for that matter that is struggling are to run, by bumping it up in 12.5 MHZ increments (or down, but obviously first you want up) you can usually find a stable setting and life goes on.  However I had one unit (S9) that I decided to upgrade the firmware on and thats when I found this stupid frequency issue.  They (bitmain) even warns you about it before you do it, but since it was one unit I went with it.  I don't know what the purpose of it was other than to stop or discourage people from bouncing all over the place with the frequency.  I suppose a lot of people damage their units with this option trying to squeeze every last hash out of their unit.  What they don't understand is if the frequency is too high the machine actually runs slower.  This isn't like a hyper-hash drive in a star wars movie.  This was probably bitmains way to try and curtail the trouble reports.   

They just removed the link, the old page still works; like: http://192.168.1.202/cgi-bin/minerAdvanced.cgi

The main reason they removed is is because the newer hash boards are preset per chip (one could be 512 and another 693), they wanted 1 firmware to update each time they modded something (instead of 10), and setting frequency on the newer ones does nothing but make you feel like you have some say in the matter.

[Make-A-Buck]

I don't know about auto-tune but it sucks!  You have one shot to pick your frequency and then it is locked in and that tab disappears in the settings.  The only way to try different frequency to find the optimal setting is to re-flash the entire device and choose another frequency and subsequently get re-locked in.  I myself am a default kinda guy.  I don't overclock, or mess with the settings along as it hashes like it supposed to.  However in some cases I have found on the S9's or all antminers I have used since the S2 for that matter that is struggling are to run, by bumping it up in 12.5 MHZ increments (or down, but obviously first you want up) you can usually find a stable setting and life goes on.  However I had one unit (S9) that I decided to upgrade the firmware on and thats when I found this stupid frequency issue.  They (bitmain) even warns you about it before you do it, but since it was one unit I went with it.  I don't know what the purpose of it was other than to stop or discourage people from bouncing all over the place with the frequency.  I suppose a lot of people damage their units with this option trying to squeeze every last hash out of their unit.  What they don't understand is if the frequency is too high the machine actually runs slower.  This isn't like a hyper-hash drive in a star wars movie.  This was probably bitmains way to try and curtail the trouble reports.   

They just removed the link, the old page still works; like: http://192.168.1.202/cgi-bin/minerAdvanced.cgi

The main reason they removed is is because the newer hash boards are preset per chip (one could be 512 and another 693), they wanted 1 firmware to update each time they modded something (instead of 10), and setting frequency on the newer ones does nothing but make you feel like you have some say in the matter.

That's a nifty lil tidbit of info right there.
I just went to http://192.168.x.x/cgi-bin/minerAdvanced.cgi on my new R4 Batch 2, and sure enough, the Freq Mod page comes up.
Not going to mess with my R4 B2 tho, it's been hashing like a champ since day 1 (which was only 4 days ago). Like they say 'If it ain't broke, don't fix it.'
Thanks ComputerGenie.

[kano]

Well it's something I added to CKDB long ago to be able to change the miner reward (to put it up twice before)
Here's the message just now:

Code:

[2016-12-16 22:13:02.186+11] process_pplns(): *** block 443699 payout reward overridden, was 1278060860 now 1389667900 = 111607040 (8.7325%) more

There was also something I was waiting on this block for that really had me wondering if it was some sort of vengeance by the block Gods
I created another node to test a less modified version of 0.13.1 so I can use it on all the nodes except the main pool - to send/receive blocks to/from the fibre block relay network faster. That pretty much requires 0.13.1 to optimise it's use.
And yep, yay, that block was properly submitted and accepted on the test node so that's all good now too, I'll update all the nodes - except the main pool - in the next day to 0.13.1 without segwit

This may be a silly question and I may have missed a thread that could have explained this but here you go.  Why wouldn't you do the same update to the main pool?  

The main pool generates the work, all (main and nodes) submit blocks.
Submitting a block on a node that no one is mining to, has no chance of causing any problems if it fails.
The initial test worked fine.

I've now switched 3 of the nodes over to the 0.13.1 version.
Aside: the SG node is now a new one, no longer linode. I've also setup a us-east node so it's close to the us-east fibre relay.
The 0.13.1 has, of course, continued to work as expected on the 3 nodes with the new blocks

The work generation for core bitcoind is different now because they have put segwit in it and I'll need to do work generation testing (without segwit) before I switch that over.
Simply, it will require more testing than needed for nodes.

Everything is either 0.13.0 or 0.13.1, but 0.13.1 has a small advantage with using the fibre relay.

[kano]

Block! by jmww

[clgrissom3]

Block!

That's 3!

[hhumaidan]

  Lets get those lost blocks

[philipma1957]

block we must  3 is good but 4 is more

[kano]

Over the next few hours I'll be updating the final 3 'remote' nodes JP, DE and NL to the newer 0.13.1 hack
This also includes updating the LTS OS since that's useful to do also.
No one should see any down time.

Basically I just switch the DNS, redirect everyone on JP, DE or NL to another node (after an appropriate wait), then update ... and undo it after.
The final side effect is that everyone currently on JP, DE or NL will end up on NL (that I'll do last) and only switch back to their original node, if it wasn't NL, when they reconnect themselves; since the final redirect will be back to NL from DE and everyone will be on DE by that stage.

No one mining to any other nodes, or the main pool, will be affected at all.

I'll post again when it's all complete.

Completed.
I left everyone from NL on DE coz it avoided a 3rd redirect back to NL of everyone.
90% of NL+DE are on DE, so better not to send everyone back to NL again.

Also there were about 15 users (out of 81) who didn't redirect from DE to NL when I did the DE update.
My guess would be they were connected direct via IP address (and thus cgminer wouldn't redirect)

Edit: So the final step, over the next week leading up to xmas is testing and then updating the main pool node to 0.13.1
All other nodes are now latest LTS and 0.13.1 core (with segwit disabled)

[d57heinz]

@kano

ok ill request this one more and i promise ill never request it again.  Can you please add 100, 200, 500 block stats to this page. https://www.kano.is/index.php?k=luck  If no can you please tell me why you will not?

 Looking at those stats will help me(us) to compare previous low luck periods. Not that it matters i know but just for fun;) if nothing more.  If you wont/cant get that done with all your work your doing,  THen can someone here please direct me on how to setup my own graph based off the numbers from the blocks page.  Just point me in the right direction. I like a good challenge.  TIA

Best Regards
d57heinz

[hhumaidan]

Block by cashdc!

[zOU]

Completed.
I left everyone from NL on DE coz it avoided a 3rd redirect back to NL of everyone.
90% of NL+DE are on DE, so better not to send everyone back to NL again.

Also there were about 15 users (out of 81) who didn't redirect from DE to NL when I did the DE update.
My guess would be they were connected direct via IP address (and thus cgminer wouldn't redirect)

Edit: So the final step, over the next week leading up to xmas is testing and then updating the main pool node to 0.13.1
All other nodes are now latest LTS and 0.13.1 core (with segwit disabled)

My 3 miners are on DE with NL as a backup.

I just checked and I saw no share submitted to the NL node, so I guess they didn't fail back (although they're set to mine on de.kano.is not on direct IP)

They're currently happily mining on DE.

[HagssFIN]

I'm back at the NL after trying my luck for few days at solo.ckpool.

[clgrissom3]

Block by pavlidmavrom with 20.18TH/s!  This is his 2nd Kano block and it's just in time to be our 2nd of the day!  

[philipma1957]

more blocks please!!!

and the five block % is at:

Description------Time----MeanTx%-----Diff%-----Mean%-----CDF[Erl]---Luck%-----?PPS%

Last 5 Blocks   2.5days   104.8%   85.22%   85.22%   0.4220   117.35%   121.92

[kano]

I'll add another 1BTC onto the miner reward for 444050 also (will show up when I tell CKDB to regenerate the payout soon)

... also since that block was 42% Diff

... last time I actually added 1.1160704 BTC to the miner reward, but this time just 1BTC

Edit: Done

Below is first expiring the generated payout, then the override when regenerating it

Code:

[2016-12-19 11:14:13.487+11] 1.ok.expire payout 2657 block 444050 reward 1313531033 status 'G'
[2016-12-19 11:17:34.642+11] process_pplns(): *** block 444050 payout reward overridden, was 1313531033 now 1413531033 = 100000000 (7.6131%) more

[philipma1957]

I'll add another 1BTC onto the miner reward for 444050 also (will show up when I tell CKDB to regenerate the payout soon)

... also since that block was 42% Diff

... last time I actually added 1.1160704 BTC to the miner reward, but this time just 1BTC

nice of you!

[firetreeactual]

Way cool. Best block reward in a long time.