This is an issue I think I started having around the first of the year.
I’m using the same computer for Plotting and Farming. It’s running Windows 10, and I’m using the latest gui version of the Chia Blockchain software for farming and plotting. I reboot when adding a farming drive, as they are all sata connected. When I restart the gui, it starts farming immediately like it should. If I start 1 plotting queue everything works great. Normally I do a simple manual staggering of the queues, so I’ll wait 30 to 60 minutes and start the 2nd one, and so on. Through trial and error I found that the computer can nicely handle 4 plotting queues each using a different ssd at once, without it affecting farming.
Sometimes, not all the time, when I start the 2nd queue farming will get stuck. Sometimes it gets stuck in minutes, other times it might take an hour. I’m not a command line person, so I have to restart the gui when this happens to get it farming again. It seems the only way I can get the plotting going properly without it affecting the farmer is if I let the computer farm without plotting for an extended period of time. Usually overnight. Then in the morning I can start 4 staggered queues normally without a problem. And the machine will work well, as expected until the new farm drive is full.
What’s going on here? I didn’t really pay attention to the first time it happened, because I thought it might be a fluke, because this behavior doesn’t happen every time. Sometimes I can just start plotting with multiple staggered queues immediately no problem. But it keeps happening, maybe every other time I reboot. Maybe started around the first of the year. Very annoying. If anyone has any advice, I’d love to hear it. I feel like I have the computer and my network configured correctly, because sometimes everything works fine. And eve when it’s in this faulty state, but with a single plotting queue running there are seemingly no problems. I’m a bit stumped by this.
If you’d like to restart your farmer without interrupting your plotting, you can use the powershell alias listed here to make it a lot easier, just open powershell, paste the alias, then run chia start -r farmer: Adding chia.exe to your PATH on Windows for easy use
As to what’s causing this (if your logging level is set to INFO) check your logs for large blocks of Don't have challenge hash <hash here>, caching EOS. I noticed I was getting roughly 2 minute blocks of these and losing sync for those 2 minutes. After changing out my database for a backup that I made using sqlite vacuum, that stopped happening. The vacuumed database seems to have improved the performance of the node.
1 Like
A few questions:
Are you plotting to the same hard drive?
Have you reformated the SSD?
Have you checked the quality of the SSD using a program to check the hardware of the SSD?
Are you using -w in your plotting cmd line?
Are you using madman alone or in chia gui?
How many threads per plot run have you assigned and does your cpu has it?
Sorry but this is what I would check if I had same problem.
Keep me up to date?
Thanks, definitely going to use that powershell command next time this happens.
As for the vacuum Sqlite command. This may sound naïve, but I thought that (for lack of better words) the chia database was the chia database. So there could be garbage in there? I do have some not too old backups I could try this with.
Here’s a pretty good guide on making a live backup using the sqlite “vacuum into” function, you could then replace your in place db with that vacuumed backup. On windows, you would just download the sqlite tools from the sqlite website, the command should be the same, just adjust the directories accordingly.
Thanks Normanvalez,
I have 4 2.5inch SSD’s dedicated for plotting. And I actually did try reformatting them at one point as something to try. All have passed smart tests. I’ve tried starting with a different drives to see if that makes a difference. The only thing I’ve been able to tell that consistently makes a difference is if I let it sit overnight farming before starting to plot, which makes no sense to me.
I use the gui to do my plotting. What does the -w do on the command line?
I’ll need to look when I get home tonight, but I normally use the default settings for the plot.
When I first started with Chia, I was doing 8 plots at once because I thought the computer could handle it. This was a huge mistake. I had the resources and it seemed like the computer was handling it, but farming was always behind. Never got stuck mind you, just behind. Cut in half to 4 at once, when it works, farming keeps up nicely and I’ve even won some XCH.
it waits for the plotter if your are ploting more than 1 in same disk to wait until done writting to start writting next plot.
dedicate no more than 4 threads per plot on each plot command. you must have at least 10 cores which is 2x - 20 theads. use madmax core and not the gui. it is slower and it wont give you all the options available. go here and download: Releases · stotiks/chia-plotter · GitHub scroll down and look for assets and click windows 64 and download and create a folder just for this software.
if you have more questions hit me up. this is where i would start. bye