how could you get so many disks to connect to a single PC? Can share detail to me? It’s a bit off topic but I want to know 
No, it won’t help, it may even get thing worse. Reason is you have same CPU, so if you break down to so many farmers, I think they will fight each other for the same resource.
My practice is I put 1 plot on day 1, get it fully plotted, then after first round of course it will replot; then maybe another replot. Once I see my first plot no more replot for a while, only then I’ll add the second plot.
But each of my disk is 2TB, so it’s different to yours, you have so much more smaller SSD disks.
here is the info on the ssds, they are not cheapo ssds
Here is the brand of the 400GB ssds
SAS Enterprise SSD HGST HUSMl4040ASS600 400GB Endurance 7.3PB
and these if i can find them HGST HUSMl4040BSS600 400GB Endurance 35PB
Here is the brand of the 200GB ssds
SAS Enterprise SSD HGST HUSMl4020ASS600 200GB
are these not good enough to work??
or is problem, i have to many? ii so whats the max limit i can use?
Why you think they’re not good enough. Let’s have a look in another angle.
Assume, just my assumption. That the replot cycle from day 0 is: 1 day, 2 day, 4 day, 8 day, 16 day, 32 day, 64 day, 128 day, 256 day, so on…
You have 18 disks, you put all of them in plot from day 0. As my calculation above, each disk, even only 400GB, requires 14400 minutes to complete (they started all at once, they will finish plotting (or replotting) all at once), which means 10 days. So until the replot cycle window reach more than 10 days, you’ll have to start next round of replot and after it finishes.
If you only put 1 plot, you will need 800 minutes for the first round, which means only 13.3 hours. So in this case, it will trigger the report after 8.7 hours.
If you just add your plot slowly, let’s say during the replot cycle window of 16 days, all your 18 disks are added as per the math above, you won’t have to replot right after the finish.
I think the bottle neck could be at the IO BW of your card, but we can only know that when none of your sectors are NOT expired, which can happen even we close this 3g testnet.
The game is a little bit complicated to farm subspace effectively.
thanks for explaining that way so i can understand, so as it seems mine will never ever be done re plotting, thats why i am seeing this loop. so i for sure will not be plotting on that pc. it will be just a replotter/farmer.
From what I’m observing on time to complete a sector, replot does the same job as plot. Maybe not 100%, but at least 80%. So no point to move the plot to other PC.
Replotting performance is self-restricted for successful farming, you can increase number of replotting threads to be the same as number of plotting threads if that doesn’t break farming for you.
SSDs you have are fine, I am using the absolute cheapest SATA SSDs I could find (56$ for 2TB) and they work perfectly fine too. High endurance enterprise SSDs are not necessary at all unless you can find them for dirt cheap used, I bet most non-defective SSDs you could find would work just fine.
The issue described in this thread is with plotting/replotting, which are both the same algorithm that doesn’t care about I/O bandwidth, you can plot on HDD without any issues, I’m not sure why you decided that would be a problem.
Yes you’re right. Sorry, I mean to say about a possible bottle neck with this card during farming (audit process), not plotting since plotting or replotting happens only at a disk at any specific time by default.
you keep telling me it replots 1 at a time, then explain this screenshot of what i been complaining about, all min hit is seconds .
is this just a bug on my end?
I think it’s one at a time - per farm. (or perhaps I should say per plot?)
As already explained above: they are all queued, it doesn’t mean they are all actively replotting. Either way I see no reason for you to care how it actually does replotting as long as it works 
If you want to see more details open something like htop and enable custom thread names, you’ll see that there are many, but only replotting threads belonging to one farm are active at any time.
@sparks60 all your farms start at same time; they all finish at same time. To prove what I say, you should look at the log lines where a new sector is plotted or replotted. It’s always almost exact the same time a new sector is plotted or replotted, based on your CPU’s power on that PC. Then, you will notice that the ball must roll over all remaining farms before it work on the first farm again.
So this proves that at a time, by default, the CPU does the plot/replot for ONE SECTOR at one farm only. When it finishes, it move to the next farm and so on. Sector index can be random. I’m not sure why but I remember I read some where that there’re 3 status of a sector.
Active: can be farmed, no need to replot
Replotting required: can be farmed, need to replot asap
expired: cannot be farm, sector is completely invalid to claim reward
@nazar-pc sorry could you confirm my understanding above?
Yes, that is correct
You’ll only see the simultaneous replotting lines in the log when it starts, at 0.00%. As Nazar says, they’re just being queued. But after that, you’ll see that it alternates between them when it comes to actually completing each sector. If they were really all being worked on at once, you’d see them all being completed at roughly similar times, but that doesn’t happen.