Does anyone know if the Orange Pi 4 would complain in some way or other if it was freed of the microphone?
Polli has taken one for the team... I'll let you know if I find that something is not working any longer.  I'm happy to announce that Polli is happily hashing without the microphone on the Orange Pi 4 !  Dang, why would you deny me all those conversations we are recording?! |
|
|
|
I used -brd_ocp 95 rather than -brd_ocp 100 because that seems to be the maximum setting possible in the GUI of the Full Package Apollo. It reports that it is using 266W with that setting.
Would there be anything wrong with cranking it up to -brd_ocp 100 ?
Not if you're using an external power supply. Limitation is there for the stock 200W PSU since we need overhead for the controller and fan power. Just watch your temps, running at 100% board power with high temps (over 70-75c) will significantly reduce the lifespan. |
|
|
|
No; you could basically turn off the fan completely by setting these parameters for instance. This should set the fan to 1% speed regardless of temperature. ./apollo-miner -fan_pwm_low 1 -fan_pwm_hi 1 <rest of parameters>
Here is my understanding of those 4 parameters. They basically set a 2-point fan curve. I used the default values as specified above.  Thanks for the explanations, I will have to spend some time on trying to understand and implement that. I thought the Fan temp settings in the Full Package Apollo would be using a temp curve similar to the one you showed above, except I imagined it actually curved. But when I set the Full Package Apollo Fan temp settings to 70c / 90c, while keeping the actual miner temperature below 60c with the added fan, then the stock fan continues to run at around 1800rpm. See screenshot. That's why I thought the stock ASIC fan cannot be turned lower than about 1800rpm.  Some fans have a minimum RPM they won't go below, it depends on the fan revision/vendor. We had to source like 3-4 fan vendors during the pandemic all with slightly different fan controllers. You can test this with the -fan_pwm_hi 1 parameter. That will force the PMW controller to not output anything and make the fan either not spin or spin at its lowest speed possible. |
|
|
|
Fan curves work through USB as well. Its the same parameters we use for the full node miner as well just didn't document them in git. For a full list just type ./apollo-miner -h FYI we are working on completely revamped support/guides/faq section on our site. These threads are obviously getting out of control, and do apologize for all the digging new users have to go through! |
|
|
|
The unit started up just fine on reboot, however the background was changed from the Apollo background to an Ubuntu desktop background, and my miner for the full unit was no longer showing up in the dashboard as a miner (the 2 standard units were still showing up and mining just fine and mining on boot).
Interesting that the 2 standard units were still mining after the Ubuntu update, I wonder how that works. The problem with the upstream ubuntu updates is that they mess with the custom bootloader we need to load to control the internal hashboard. In hindsight I should have just controlled the internal hashboard through USB as well and it would have saved a lot of headaches for me and everyone else. Since the external hashboards are through USB these kernel updates dont mess with the functionality there. Good news is we finally isolated the probablmatic updates and will freeze those updates in the next image release. This *should* finally fix the updates bricking the full node. |
|
|
|
Any updates for the LN support guys? Been a long time coming  We are still building out the app ecosystem, once this is done the apps will come fast. Also a reminder to everyone there is nothing stopping you from installing the LND apps today and pairing your node to LND. The UI will do the same thing and just install the publicly available apps in a one step process for beginners. |
|
|
|
Node has stopped running. Appears to be due to storage filling up since it's only a 500gb drive. If I'm not mistaken, the solution is to simply obtain a larger capacity sd card and insert into the bottom of the node unit, yes?
Simplest solution is to enable bitcoind pruning (which is limiting number of blocks to certain value, basically discarding old blocks, so hard drive don't overspill). But Futurebit has not got around to support their users properly - this is nowhere to be found in the settings. They knew the day is coming but did nothing, now all early batch users with 512GB drives running with crashed bitcoin nodes, not providing any benefit to Bitcoin blockchain p2p network, which was heavily advertised during sales. You are left out, you either have to configure pruning yourself in bitcoind config files inside Apollo, or purchase 1TB NVMe drive, brand new for around 70 USD (in the UK where I live) last time I checked. So I: 1. Swap the internal drive, which is more annoyance than I'd prefer to deal with.. 2. Attempt to modify the software, which is more annoyance than I'd prefer to deal with.. or 3. Ignore the situation and lose out on the money and purpose for buying a node unit to begin with? Yes, exactly, that's all options you have. As Sledge0001 mentioned, buying new 1TB NVMe drive may be easiest option, but it costs money. Second great alternative option would be Futurebit adding pruning, so you can lose out 20GB of old blocks and still serve 490GB of newer blocks. This interferes with planned in the future Lightning Node support, but LN support is nowhere at the moment. But unfortunately, because Futurebit did nothing to prevent this, we now arrived at option no. 3, where ALL nodes from 512GB users just crashed and no one is serving any blocks to Bitcoin network, and no one (apart from few here in this thread) even knows why, because there was no announcement or warning to users, no newsletter, mailing, nothing. Every node just dead. Edit: Before someone mentions this issue, to my knowledge pruned node are seeding blocks, and they are considered as "full nodes": https://bitcoin.stackexchange.com/questions/72617/how-can-a-pruned-node-be-classed-a-full-node-without-the-full-blockchain says: Since, they don't keep older block data, only thing that they can't do is to return older block data to other nodes. They still keep latest block data for reogranization, which they can share with other nodes Amused. Just got an email from FutureBit perfectly explaining how to upgrade the drive. At your own cost, of course. No pruning in sight. And I received no e-mail whatsoever, so I assume it was you who contacted them, and another thousand of users get nothing. Everyone that ordered a Batch 1 or 2 should have received the email. I will not advocate or push an update for pruned nodes, a pruned node is the same as a dead node in my eyes. As has been discussed extensively since the beginning, we purposely chose a 500GB drive knowing people would have to swap the drive out in 1-2 years because we knew prices would be a fraction of what we would have had to charge for Batch 1-2 devices with a 1TB drive and that was the right choice then and is the right choice now. We would have had to charge an extra 150 at the time for the 1TB drive, 1TB drives are now 75 and will end up costing you less than 50 if you sell your 500GB drive. I think this is the better decision all around, and swapping out the drive and setting it up takes 5 minutes. |
|
|
|
Quick question:
How can I keep screen enabled and emulate a monitor connected, when there isn't one connected via HDMI?
Right now I resort to HDMI dummy plug so Orange Pi 4 thinks there is a monitor connected. Autologin is enabled in Ubuntu user settings and Gnome session starts nicely. But without monitor/dummy plug connected, nothing starts and x11vnc won't let me login remotely.
I am using original Futurebit image with Armbian system.
jstefanop, any update on this please? Should I write to futurebit e-mail? This is a limitation of the hardware/graphics driver. The HDMI port /GUI won't activate unless there is a monitor powered on and connected to the HDMI port (or a dummy plug like your using). |
|
|
|
Drive Almost Full - 456.74GB out of 457.45GB used ... what happens next.
Anyone have any idea what happens the drive finally is full on the miner? I've been watching the drive space dwindle now for the last couple of weeks and I'm pretty sure that sometime between Sunday and Wednesday of this week it's finally going to happen. So that said what will happen, will the units, die, will there be an automatic FIFO delete?
Is there a good idea of what can be deleted via SSH'ing into the box, I don't use the node but it has run and I'm sure there is data in there that could free up some space if someone can tell me what to delete that won't cripple the miner.
Looking for suggestions.
Thanks
futurebit@futurebit-btc:/$ df
Filesystem 1K-blocks Used Available Use% Mounted on
udev 1902880 0 1902880 0% /dev
tmpfs 395612 15996 379616 5% /run
/dev/mmcblk1p1 14899824 5905236 8815336 41% /
tmpfs 1978056 0 1978056 0% /dev/shm
tmpfs 5120 4 5116 1% /run/lock
tmpfs 1978056 0 1978056 0% /sys/fs/cgroup
tmpfs 1978056 8 1978048 1% /tmp
/dev/zram1 49584 10632 35368 24% /var/log
/dev/nvme0n1p1 479669928 478929108 236052 100% /media/nvme
tmpfs 395608 8 395600 1% /run/user/111
tmpfs 395608 8 395600 1% /run/user/1000
futurebit@futurebit-btc:/$
Best bet is to do like lots of us have done and upgrade to a 1TB SSD. But to answer your question, I have no idea what will happen once you reach max capacity. I'd imagine you'll just stop downloading new blocks Yes all Batch 1-2 Apollos will need to upgrade their SSD drive soon. If you dont your node will stop working. As we mentioned from the beginning we did this on purpose since 1TB drives were nearly 200 USD then. You can get good 1TB drives for around 75 USD now, and a 1TB drive will last at least 4 years even if every block is full going forward. |
|
|
|
Hello,
I recently purchased the FutureBit Apollo BTC full node. It was running fine. Then I updated the Ubuntu software (at the request of a pop-up that appeared on the dashboard screen). Now the miner will not work correctly. When it restarted following the software update, the dashboard would not load. I tried restarting/unplugging the miner. Now when I turn it on, no light at all appears on the front of the miner. When I turn it on, the fan runs at the high intensity (not eco level) but nothing happens. I can't remote into the dashboard.
I have no idea what to do to get it going again. Any advice? Is there a way to do a hard restart/full reset?
Thank you!
Oh no, bricking Apollo software on update is back again.  It's not unlikely that Dagwood7 either did a full distro upgrade to 22.04 LTS or that the user got a unit that didn't yet have the latest Futurebit OS on it. It would be great to get confirmation / clarification; I bet also for jstefanop and his team.  They are doing the distribution update to 22.04. We need to disable this prompt in future image release as only 20.04 system and security updates are supported. If you do the major distribution update it will still brick the system and you'll need to reflash. Only perform the normal sudo apt update/upgrade on the system. |
|
|
|
Hey folks- longtime reader, rarely calling.
I’ve been running 6 apollos (one full node, 5 standard) with 6 more on the way (same as previous, 1 node, 5 standard).
Since I’ve been running them, I have had minimal to no problems. Are there any updates I should be pushing to increase mining?
My biggest problem, which I think is because of my ISP, is when my internet “burps” or loses connection for a minute or two, I have to manually restart the miners. They are hard wired via Ethernet to a router. No big deal… when I’m home. Is there a way to somehow access my Futurebit site when I’m not on my home network? A friend mentioned port forwarding?
I’ll preface that I have a solid understanding of network technology, but nowhere near advanced. Fortunately I learn quickly.
Im sorry if this has already been discussed, I wasn’t what to search for when trying to find answer.
If anyone could point me in the right direction or any material that can help I’d be greatly appreciated.
Thanks and happy mining!
Chickywinger
P.s. @jstefanop the usb hot wiring is working great! Thanks again
If you haven't updated since you received them now would be a good time to update your main unit to the latest image. There have been updates that have resolved the issues with needing to restart after an internet outage. |
|
|
|
|
FYI we have been getting some reports of users mistakenly doing a full distro upgrade to their system, this will still brick your OS as it pulls the full ubuntu upgrade to 22.04, which our board is not designed for.
The base system will still be 20.04, and you should only do apt update/upgrade...do NOT do a full dist upgrade if the OS prompts you to upgrade to 22.04.
The fix we implemented is to only do base system updates/security updates to the current 20.04 and 5.15.x kernel which the Apollo will remain on for the foreseeable future.
|
|
|
|
satosato42 You brought some valid points. Perhaps jstefanop should address some of your concerns. Regarding having to create forum account - to get support, you can always e-mail them, no need to participate in free discussion here if you don't want to. And if we were to move this to Telegram/Discord, account creation is required there also. But still... Having " Official FutureBit Apollo BTC Software/Image and Support thread" here and its not officially listed on Futurebit website? How people should know about this official thread? Last point I wanted to address: Mega has been involved in quite a few sketchy things in the past and so downloading files from their site in not my preferred method of choice for getting essential files that a manufacturer/company should be providing on their site, self hosted.
Mega has been involved in what sketchy things? One can host files anywhere on the planet, as long as hash matches, your file is 100% safe. jstefanop provides links to download together with hashes to check. You must check if file matches the hash, after you download it. Twitter, email, and this forum thread have been good support avenues so far and until the futurebit community gets into the 100s of thousands of people (which we hope!) there is no need to dedicated more resources to support currently. Company is run very lean and thats why we have survived 3 bear markets so far and can offer our products with very low margins. With that being said product support will be revamped in the coming months with a dedicated support site that has most of the info in this thread + guides + faqs etc in a clean easy to use format. We use mega because its the most crypto friendly host, sure they probably host more "sketchy" things on the internet than most but thats not Mega's fault, as long as your cross checking the checksums I post here and on mega and only downloading from links on here or futurebit.io your downloading the official images. I will start signing images for added security though. |
|
|
|
Has anyone used the NODE Configuration to set up ledger live access? If so, would you please share how you configured the node from the settings file?
'
Have been trying and can't seem to get it working.....asked a question about it on a recent post, got a response from john but it didn't work. Prior to the new MCU1 image flash, I was able to remotely connect sparrow to the node with and without TOR enabled (per configuration settings in a previous reply). Post-flash I'm unable to connect wallet via RPC or by SSH with puTTY. I've tried every combination of username (futurebit & futurebitbtc) and password (default futurebit123, custom PW created in initial setup and the Bitcoin RPC password as shown in the settings/node config) with no luck. I suspect it may have something to do with a username/password issue during initial setup because when I Log-Out of the dashboard or the OS, I'm immediately unable to log back in using any of the possible password options. Seemingly the only way around this is to power off and restart. I've reflashed image and setup 4x and keep running into these issues. Any thoughts? The new image resets the futurebit user password to the password you set in your dashboard setup. There is also a known bug if you use a special character in your password, so please use just alphanumeric password until we fix this bug. |
|
|
|
Ran the update this morning, which is now stuck at "Hold on, page will reload in a few seconds..." It's been at that stage for over 30 minutes. What should I do if anything? Also I am just now reading about flashing the drive, which I know nothing about. I think my Apollo's are batch 3.
Edit: refreshed page, update seems to be complete. However I'm unable to move forward with the flash process as I cannot get Etcher to run even when setting permission to do so.
Hope to still get the team to help with updating the flash. The Apollo won't open any AppImage files, even if they are set correctly to do so. Double clicking to run the program only results in a spinning wheel for a number of seconds, and then nothing.
Satayasat- while perhaps obvious to most, those of us less technically inclined read the MCU Flash Instructions verbatim and ran into the same problem: Apollo wont run the balena-etcher-electron-1.7.9-linux-ia32 file. Power down the Apollo, remove the microSD card and perfom the flash instructions on a different computer. Re-install, power on, and re-configure passwords and settings as on initial device setup. Yea I should update the instruction and specify you need a separate computer and you need to physically flash the SD card by removing it from the apollo so its clear... |
|
|
|
I must say that $300 for adding this SBC and SSD is steep. The 'hashboard only' option can be acquired for $524.99 right now, whereas the 'full unit' runs you $824.99. Buying an Orange Pi 4 and the exact same SSD right now, runs you pretty exactly $200, so it's a 50% premium. I'm not sure if you can just buy those yourself and mount them into a 'hashboard only' unit, or if there are missing mounting spots or holes on those. But from pictures, they look exactly the same. Just throwing an idea out there.
Just saw this thread, thanks for the honest review. I do want to provide transparency whenever I can especially when it comes to comments like above when most people have no idea costs/development involved with a product like this. We call it "our" controller because it truly is, and there is nothing else on the market that even comes close to its performance for a Pi sized computer and thats mostly from the innovation with our custom SSD controller. AFAIK its the only SBC on the market that has 1TB R/W speeds which is why its so much faster than a PI when syncing. Sure the base board is off the shelf (we have custom changes to this as well), but we put in a lot of work to the overall system with the SSD components. Pricing, here is the our overall bulk costs we paid for Batch 2 and 3 for these components SBC:95 SSD:90 SD Card: 9 Custom SSD card: 19 Heatsinks/Fans/Wire harnesses and other small components: 25 Thats 240 and does not include imports costs/shipping costs/assembly/R&D etc etc. When all that is factored in we actually make a net loss on the full node units in relation to its price premium. The little profit we make is from the hashing unit itself. 300 is a steal for what you get, especially when compared to other full node only competitors. FYI most of the other issues especially around upgrade the base OS and security issues have been fixed in the latest image release. |
|
|
|
Excellent news! Congrats to the whole Futurebit team on the update.
I look forward to updating as soon as I backup the Home Assistant data, which I just installed on Monday and loving so far (for those interested).
Two questions from me:
- being a Batch 3 owner, will there be future updates/features that come to the new MCU2 and not to MCU1? Hope I'm on a 'retired' or 'legacy' hw platform after only a few months.
- after reboot, at the login screen, is there a reason virtual keyboard is disabled/not loading? I share a keyboard between my workstation and futurebit. Was hoping to use the virtual key board to login, but doesn't appear to work. Was this intentional?
Both MCUs are identical and will be supported with any software updates, the new version just has different components that are more readily available and needs a different kernel to run so thats the reason for two versions. Not sure about the virtual keyboard, is this occurring with the new image? |
|
|
|
New image is out! Batch 4 units come with a new MCU that is no longer compatible with the original MCU firmware so there are now two separate image files depending on the MCU you have. For most here you're on MCU1 so make sure you download that version. This image fixes the critical system update issue and is the largest upgrade since release. If you are a Batch 1-3 owner its highly recommended you flash your SD card to this image, and you will finally be free to update the system whenever it prompts you to or do apt upgrade/update without bricking your system! All Batch 4 units shipping now already come with the latest firmware installed at factory so no need to flash to this one. Folder for all Apollo BTC Image Releases: https://mega.nz/folder/ZlxnwYZb#ZN77cDAcXa6dN_2QKSdT_QThere are now two versions of the Apollo MCU, and firmware images are not compatible with each other. Please read flash instructions below before proceeding to flash your unit!https://mega.nz/folder/ZlxnwYZb#ZN77cDAcXa6dN_2QKSdT_Q/file/Fg4AzABDRelease 7/25/22eaac461245fd4bc2a410f2dc5ebc861add6aa5618405c5998f780247999570b3 apollo-btc-mcu1_250722.img.xz 4d46819c9af6e6050b5d7c395da33aa4c10616bd46261a4fc4a3482e31ed36be apollo-btc-mcu2_250722.img.xz -Updated Apollo UI to latest 0.3.4 -Updated system to latest 5.15.xx kernel and package/security updates -FutureBit user login password is now updated to dashboard password after user completes dashboard setup for added security -Bitcoin core folder on nvme SSD drive now symlinks to standard location (~/.bitcoin) -Fixed critical issue preventing system from updating -System level fixes/tweaks for stability |
|
|
|
trying to connect my node to ledger live....will this work ?
has anyone tried and got it working ?
Yes, with the new update all you have to do is add the line rpcallowip=youripaddress in the settings page of your node and save. IP address has to be the local IP address of your computer you have ledger live installed. You then can enable the connect full node settings in ledger live using your nodes ip address and its RPC credentials (username futurebit RPC password displayed in settings). Obviously both the node and your computer have to be on your local network for this to work. That should do it and allow you to broadcast your Bitcoin txs through your own node via your ledger hardware wallet! |
|
|
|
I am updating via the UI.
Been sitting on the 'Hold on, page will reload in few seconds...' for an hour.
Received the Slushpool Monitoring: 1 | 0 | -1 | 0 we all dread.
Is 'in few seconds...' a euphemism for something or should I conclude that my update has failed?
Restart close the page and login again. Worked like a charm, I should have checked the worker on Slushpool before posting as it has been active for 35 minutes. Yea if its stuck on the page will reload for more than 10 min you can just refresh the page and most likely the update has been completed. |
|
|
|
|
Show Posts
