You could try mine, but try it on just one to make sure it works for you first.

Jason

I’m on the road at the moment, but I am willing to try to help via team viewer at some point if you would like.

Have you tried ssh or just http?

Correct, no Nicehash. I have spent a lot of time on that but now need to work with NH engineering. The action is on my side to do that at this point.

As far as the API goes, HiveOS is correct. On the box I actually insert the frequencies into the API for the web interface. Exporting that off box via the normal API is probably going to be a no. I will add it to the queue and see what I can do there, but the API handling inside cgminer is a mess of spaghetti.

Thank you,

Jason

Folk,

A question has come up on how to tune Z9 and Z9mini with the per hashboard feature. Here is a small writeup(copied in the original post now) that hopefully will explain:

To tune for best frequency

1) Set global frequency to 650. Set each other frequency to "Use Global".
2) Observe your boards for at least 10 minutes, if any drop out, on the next cycle decrease their frequency by 1 step in the pulldown.
3) Repeat this cycle until you have a stable minimum frequency for your boards.
4) To start the 'upwards tuning', choose one board (#1 for example) and increase it's frequency by 1 step.
5) Observe the board for at least 10 minutes to see if it is stable. If it is, repeat #4. If it is not, drop down 1 step and move to the next board.

Going through this process (which hopefully makes sense), you should be able to tune your units for your individual system maximums. It's best to do this with the fans set to 100% so you can remove thermals from the equation -- after which, set the fans back however you want them and see if your boards stay stable. If they do, good, of they don't, eithee decrease freqeuncy by 1 step or increase fan.

Please do note that sometimes a lower frequency produces a higher hash rate, so be mindful of that as you select for your environment.

Thank you,

Jason

htautosjay - it will vary to be honest. With the per-hashboard feature, I've been able to get +5% out of batch2s. That isn't 16k on most, but is an additional 5% from baseline. Tuning takes some work though. Dev fees are disabled until 12/1, so you are welcome to try it on a mini and see.

To tune:

1) Set global frequency to 650. Set each other frequency to "Use Global".
2) Observe your boards for at least 10 minutes, if any drop out, on the next cycle decrease their frequency by 1 step in the pulldown.
3) Repeat this cycle until you have a stable minimum frequency for your boards.
4) To start the 'upwards tuning', choose one board (#1 for example) and increase it's frequency by 1 step.
5) Observe the board for at least 10 minutes to see if it is stable. If it is, repeat #4. If it is not, drop down 1 step and move to the next board.

Going through this process (which hopefully makes sense), you should be able to tune your units for your individual system maximums. It's best to do this with the fans set to 100% so you can remove thermals from the equation -- after which, set the fans back however you want them and see if your boards stay stable. If they do, good, of they don't, eithee decrease freqeuncy by 1 step or increase fan.

Good luck!

Jason

chipless, trying to help with this response regardless what you may think....

I have not tried chipless' firmware, but the default http login is root/root and the default ssh login is root/admin. If the password is changed via the web interface, then the ssh and web logins are synchronized.

I disagree with the comments on this being network issues (ignoring IP changes) or browser caching -- that simply doesn't happen for HTTP Authorization requests.

If you do a _full_ factory reset using the V9 procedure, the MAC address on the unit will change and your router will assign a new IP address (Incidentally, I've had the instructions on this recovery at the bottom of my first post on the firmware thread from day one for worst-case scenarios).

Unless something has changed since the version of chipless' firmware I looked at, I'm going to agree with him and say I cannot see how this would happen. If you install firmware with the 'keep settings' unchecked, all configuration is removed except for the MAC address (which means IP address is likely to stay the same, even if using DHCP; this is part of my disagreement on the network issue comments).

If we assume that the passwords are out of whack for whatever reason, then yes, the V9->Z9 recovery process will work, 100%. Personally, I'd try to look at the ssh login and see if root/admin works (and root/root, in case something wacky has happened).

@chipless - in this scenario, I've worked with the users via teamviewer to understand and resolve the issues. That may be an option if you can get one of these users to work with you. Does not take long and goes a long way towards customer satisfaction as well as knowledge gained.

@Kunolyb -- I do not completely understand your response. smarty's firmware is for the Z9 and chipless' firmware is for the Z9Mini. How and why would you have upgraded from smarty's to this one?

Thank you,

Jason

chipless,

In the future, please try to refrain from posting this much of your log here; Only about 6 lines of that log were needed.

Two things can cause this: 1) network problems, 2) two cgminer processes runing.

1) In your case, the log clearly tells you there were communication errors with the pool. That is a network issue, caveat'd that it could be #2, so check that first. Otherwise, the FIFO queue full message is exactly that. Can’t submit work If the stratum cannot be communicated with.

2) This also happens if two cgminer processes are running which can happen due to a race condition in the factory firmware between /sbin/monitorcg and any operation which calls "/etc/init.d/cgminer.sh restart", which happens when you change frequencies. Solution is just to reboot. Or, for someone such as yourself, observe two are running, kill them via ssh, let one start back up.

3) As far as the fan going high, someone such as yourself should be completely aware of the startup sequence now... it goes low on init, then high until it is fully online, then the temp fan curve control takes over.

4) Please provide your 'Support ID' in the future, located on the Summary page. Using that I can help you better.

Net-net, looks like it is working as expected.

Jason

Yeah, you bring up an interesting thing that is slightly complicated in terms of component interactions. First, things like minerstat (and awesomeminer) do not handle pool switching as elegantly as they could with cgminer. What I mean by that is the pool can be changed (at least for pre-defined pools) on the fly without restarting things. In general that is not how they work though (awesomeminer does with some right click combination). The second is I can envision some situations where a utility like minerstat might trigger a tamper mechanism (purposefully vague here). I will have to think on how to resolve that interaction, but I wouldn't worry about it at the moment.

If cgminer is restarted by minerstat, then yes, the timer would essentially get restarted each time so that initial devfee would hit each time.

It is in my plan to handle things like this on the server side at some point in the future -- in other words, the timer for dev fees would not get reset when cgminer or the whole miner was restarted. To be honest, I only planned to tackle that piece if the usage scales to a point that necessitates that....  it would mitigate this concern completely, however (while enabling some other capabilities that folk might appreciate).

To your question on a non-devfee version, I am not taking additional "paid" customers on until I can get the previous set upgraded. They have been waiting a long time to get the features that were already in the dev-fee version. I will update the thread when that door for sales opens again.

I know this isn't an ideal answer to your situation and question, but I definitely appreciate you bringing it up. It is a condition I had thought about but forgot to document.

As an aside, when I first started my mining adventures, I did the pool hopping thing, which then grew to "I'll just get more miners and split across the pools"... and then got crazier from there. In the end, my _personal_ experience is that pool hopping cost more than it gained. There are a number of reasons for this including the "difficulty" with the pool gets reset when cgminer restarts, which means your share submission relative to difficulty will take a while to ramp back up. I'm not attempting to dissuade you, just share my similar experience.

Thank you,

Jason

Ah, ok. So the flightsheet doesnt' let you control individual frequencies, but once you set the pool config there, you can tune the frequencies on the miner. Got it. Yeah, that makes sense.

Thank you!

Jason

Mini support is now available in the Z Series firmware. Please see my other thread for details.

Thanks!

Jason

Awesome! Can you tell me more on this? What is flightsheet? Is this something other users of hive should be able to do easily?

Jason

I cannot speak to HiveOS other than I think it will work with it, however HiveOS would have to add support for board frequencies.

Jason

Ah.. sorry, thank you for catching that. It is fixed.

Jason

Version 2.1 is now available for Z9, Mini, and Paid users; please see the original post for information. I will update the post with more details in the next few hours.

At this point in time, I am not selling additional licenses -- I need to get these images out there and the regular paid users updated before I expand; in the interim, I welcome you to use the dev-fee supported version to ensure that this firmware meets your needs.

Thank you,

Jason

freq per chain will be released tonight in my primary thread for Z9s -- mini support is going to be available with the release. c.f. https://bitcointalk.org/index.php?topic=5036968.0

voltage will be coming in the future if viable.

Jason

As an aside, I have 2x $55 and 2x $90 bitmain coupons available that are good through the 30th. first come first serve for free *IF* you are going to use them.

thank you,

jason

Truth; I am not the one who assumed the previous response was an attack, however. And smarty has not been a target; he did his work, acknowledge mine, and I his. that was that. stop comparing yourself to others.

Mine shows 4th just fine, thanks-- this has been commented on by a few so far. And to be fair, I didn't know the answer until your response; at which point I took the time to figure it out myself. So stop assuming, since you are flat out wrong each time you have assumed so far.

If you call that talking down, fine.  Don't threaten me though. rip mine apart, good luck. You'll need help to get through it since you cannot even figure out what the simple shell expansions mean and assumed that shell matches and grouping (example ${foo#bar}) meant correlation in variables inside the firmware. Honestly, I welcome the competition. Please, rip anything past my initial release apart. Let me know when you figure it out because I welcome the awareness on any mistake I made. Don't assume me to be a junior idiot. I can acknowledge mistakes.

I actually apologized to you, took the high road, then you commented that it was a "pissing match" followed by "lame that I...". Grow some, own your mistakes and apologize yourself.

Jason

@tahoe - since I don't have everyone's email, no. But, your original link will currently work to provide you "half" of the instructions. Your link will currently provide you your license and instruction -- I just need to finish the image.

I tried to keep track of who contacted me how, but it became a little complicated.

I will release an image tomorrow night (well, late tonight now since it is 12:36AM at this point) for paid users at least. I'm confident in that infrastructure and the image now, save for one bug that won't take me long to fix.

Jason

Image will be released for Z9 paid, and mini users tomorrow evening, 11/26/2018. I've received feedback from the beta testers and will be incorporating the necessary fixes.

The delay this time is simply it was Thanksgiving and family took priority. :-)

thank you,

jason

Was not a pissing match.  Was an honest-to-God question, hence me approaching as "correct me if I'm misspeaking". And I'm glad this is working for you, honestly.

You said it was "cgminer from batch4 firmware in the batch1 image" -- What else was different? The _only_ portion of the image that actually interacts directly with the ASICs *IS* cgminer, so I'm having a hard time understanding what else could have possibly changed. Rephrased: how is what you did any different than just the batch 4 image by itself?

If it was a pissing match, I'd point out that all you did was copy one file.... which tells me there is zero reason that the latest firmware wouldn't just work on all batchX minis. The only other differences in batch1 and "batch4" as you call it from an image perspective is that batch4 added an updated libncurses (and given the opkg configuration, was likely a mistake on bitmain's part), and that the HTML UI was updated to not look for a fourth chain.... So "your" firmware will not work with 4 chains -- at least from a presentation perspective.

Jason