So I’m not sure what I am doing wrong. I’ve got 4 1tb ssds for plotting, and four 6 tb platter drives to move the data to. I’m running a i7-5930K 3.50GHz and 64gig of ram but I can’t even make this thing plot 1 plot.
The ssd and platter drives were connected to the system board, now the ssds are connected to a sata card because I wasn’t sure if I burned out my system board. (I didn’t)
The other day I started this out with trying to to load 20 plots per ssd run in sequential order not in parallel. Not even ten minutes later the chia app went non responsive and my one ssd drive disappeared. After a reboot the one drive came back, and I tried again but this time at 10 plots per ssd using 5 of 6 cores and 6750 for the ram. Again after a few minutes my drive disappeared. When I rebooted the machine this time I had to format the drive to get it to respond properly.
So before I went to bed for the night I tried it again on a different ssd and platter and only set it for 4 plots with default ram and 2 cores then went to bed. This morning it was at 8% and 10% after I got home from work. When I even moved the mouse the chia unit went no responsive and cpu usage was at 0%. This evening I tried just doing a single plot and the cpu wasn’t even working at all and the Cia app went not responding very shortly after. Crashed my ssds hard enough it took a full shut down and restart to get them to come back.
I’m not sure what I’m doing wrong here. The ssds are samsung 870 qvo 1tb with read and write speeds over 500MB/s. I’ve even tried SSD to SSD and no difference.
Yesterday i was going over the beginners guide and the one thing I tried different was that i limited my threads to 4 and 5000 for the ram. I thought this was working but when i got to it this morning the chia.exe precess had used up a large amount of memory but the cpu was only at 1% usage. I tried this again this morning and kept a watch on it. It looked like it was going well for 20 minutes then all cpu usage dropped to nothing and the ssd was at 0% usage and never started working again.
In the task manager there were 4 chia.exe procesess with the disk icon and 4 with the chia leaf icon.
Anyone have any thoughts? I have ran the samsung magician application and it says the drives are fine.
First make sure no leftover processes are running (do a reboot)
Start the GUI and wait for it to be fully synced.
Stat a single plot with default settings.
Let it completely run till the end.
See where it bugs out or gets bottle-necked.
Retrying in five minutes.
Only wrote 0 of 227324 bytes at offset 5282213684 to “W:\Plots\plot-k32-2021-05-29-13-11-93c37dae730cab7602be99315d452c9481e36ac8b5b354f702289f6d49f712d6.plot.table1.tmp” with length 5282213888. Error 1. Retrying in five minutes.
Only wrote 0 of 227324 bytes at offset 5282213684 to “W:\Plots\plot-k32-2021-05-29-13-11-93c37dae730cab7602be99315d452c9481e36ac8b5b354f702289f6d49f712d6.plot.table1.tmp” with length 5282213888. Error 1. Retrying in five minutes.
This time I get this. I think the drives are shutting down.
2021-05-29T22:12:46.503 chia.plotting.create_plots : e[32mINFO e[0m Creating 1 plots of size 32, pool public key: 83b78b3d9d5731ed219e52d8715acb21cdb4df2452fd77e004202d458f5523380373258b682b55523b30330325af04d7 farmer public key: b8a88f25bc0fc8c4a4b56e9f0abbf55f5a104dacfa3828058a2d16b461b4e813427983b8287e1c8a6d79e5addf425fc9e[0m
2021-05-29T22:12:46.511 chia.plotting.create_plots : e[32mINFO e[0m Memo: 83b78b3d9d5731ed219e52d8715acb21cdb4df2452fd77e004202d458f5523380373258b682b55523b30330325af04d7b8a88f25bc0fc8c4a4b56e9f0abbf55f5a104dacfa3828058a2d16b461b4e813427983b8287e1c8a6d79e5addf425fc941ef4e80455c748177859dad016284e0ddb4f45957d0b6ef4687d743dc780e67e[0m
2021-05-29T22:12:46.511 chia.plotting.create_plots : e[32mINFO e[0m Starting plot 1/1e[0m
Starting plotting progress into temporary dirs: W:\Plots and W:\Plots
ID: 22ef2f8e61ebc0229defa0f1714e14531ad714e3c5b7de1d6b3cb96ea86145a6
Plot size is: 32
Buffer size is: 3390MiB
Using 128 buckets
Using 2 threads of stripe size 65536
Starting phase 1/4: Forward Propagation into tmp files… Sat May 29 22:12:46 2021
Computing table 1
F1 complete, time: 215.877 seconds. CPU (104.24%) Sat May 29 22:16:22 2021
Computing table 2
Caught plotting error: Could not open W:\Plots\plot-k32-2021-05-29-22-12-22ef2f8e61ebc0229defa0f1714e14531ad714e3c5b7de1d6b3cb96ea86145a6.plot.p1.t2.sort_bucket_003.tmp: Invalid argument.
[10840] Failed to execute script chia
Traceback (most recent call last):
File “chia\cmds\chia.py”, line 81, in
File “chia\cmds\chia.py”, line 77, in main
File “click\core.py”, line 829, in call
File “click\core.py”, line 782, in main
File “click\core.py”, line 1259, in invoke
File “click\core.py”, line 1259, in invoke
File “click\core.py”, line 1066, in invoke
File “click\core.py”, line 610, in invoke
File “click\decorators.py”, line 21, in new_func
File “chia\cmds\plots.py”, line 135, in create_cmd
File “chia\plotting\create_plots.py”, line 176, in create_plots
RuntimeError: Unknown exception
so two things. 1) found a video where he made an offhanded comment as to how many threads were per core, so I lowered my thread usage. 2) I cut the amount of ram usage in half. So even with 64gigs of ram, it seems like the 3300 by default is to much.
It;s currently holding and working, see if it working in the AM.
Its not much, but it may help:
Did you uncheck the MBR feature at the Disk-Format?
Did you format the SSD to NTFS?
Did you check ports in BIOS for AHCI or RAID?
Do you use the drives as RAID?
Did you check your “addiitonal Power Setting”?
I found searching “Event 129 Storahci - Warning” that your Samsung SSD may need an Firmware update. (something called “Samsung Magician software”)
I had similar issues with a XPG8100 M.2 Nvme drive. It kept heating up and crashing. I replaced it with WD Black 850 and now everything is fine. Your plotting drives may not be suitable to handle the workloads.
Did you uncheck the MBR feature at the Disk-Format? - GPT
Did you format the SSD to NTFS? - NTFS
Did you check ports in BIOS for AHCI or RAID? - ACHI
Do you use the drives as RAID? - no
Did you check your “addiitonal Power Setting”? - No on this i didn’t think power was an issue.
I found searching “Event 129 Storahci - Warning” that your Samsung SSD may need an Firmware update. (something called “Samsung Magician software”) - Yup already had it, didn’t want to run the firmware update, cause they were working. But i did and now all four show up as
"disk #
Unknown
Not initialized
and now when i try to initialize it i get “a device which does not exist was specified”
I am not 100% on this. but it may be just more than speeds that is needed here. The robustness, build quality and heat management may come into play as well.
I have the same problem with Firecuda 510 2TB and Aorus 1TB.
Both are plotting for 24 hours and then suddenly read/write errors and “A device which does not exist was specified” error
If I try to delete volume or format in Disk Management they disappear.
After a restart they work fine for like a day.
Any ideas?
I have a Firecuda 520 1TB that I am using and it is also crashing in a similiar fashion. My OS is on that drive so it would crash and than the computer would reset. I am thinking the drive is overheating. Tomorrow I plan to swap the Firecuda for a spare WD Black 750 to see if it fixes the issue.
Not suitable and not working are 2 very different things.
Both my Firecuda 510 2TB and Aorus 1TB would plot for 24 hours, create 15 plots and then fail.
After a restart I can do the same thing again.
So i ended up taking all 4 drives back, i crashed out last night without even the chia system running and having the computer on for less than 5 minutes, 2 of the 4 drives were so hot they could barely be touched. I wasn’t fighting with them anymore. So Samsung drives, don’t buy, even if they say 500MBs
“So Samsung drives, don’t buy, even if they say 500MBs”
Statements like this are unfounded and rather dumb. Many cheap consumer grade drives can’t handle Chia plotting as the behaviour is totally different than OS boot or gaming for example.
I use Samsung 980 Pro’s which are very good drives for this sort of work.
