i am having to drop back to 1 plot at a time per pc as my internet just will not handle all this replotting., but when replotting hits, it wants do replot all i have 100% done all at same time, i dont see this ever working out for me. the replotting somehow needs to be limited to 1 replot at a time, when done move to the next. can a option be added to do this?
There is a flag to set this, by default it is 1.
–sector-encoding-concurrency <SECTOR_ENCODING_CONCURRENCY>
Defines how many sectors farmer will encode concurrently, should generally never be set to more than 1 because it will most likely result in slower plotting overall [default: 1]
So if you have plot 0 finishes plotting but replot is ongoing, plot 1 has plot ongoing, plot 2 has plot ongoing. By default, it will replot plot 0 first; then move to plot for plot 2; then plot for plot 3; and then return to plot 0 to replot, so on…
Replotting is already limited to 1 at a time by default, just like plotting is.
It will pre-download the next sector while the previous is being encoded, but it will not do anything else about it and will wait for previous sector to be encoded and written to disk before starting encoding the next one.
They are all queued, but only one makes progress at a time
i have watched them replot over and over many times, says complete, then starts over?
2023-11-07T15:00:34.709753Z INFO single_disk_farm{disk_farm_index=1}: subspace_farmer::single_disk_farm::plotting: Replotting complete
2023-11-07T19:29:59.591387Z INFO single_disk_farm{disk_farm_index=1}: subspace_farmer::single_disk_farm::plotting: Replotting complete
and you can see its reploting #1 again from last screenshot.
Replotting can “start over” if new sectors expired while you were replotting the previous set of sectors, but sector numbers will most likely be different.
2023-11-10T12:45:04.721411Z INFO single_disk_farm{disk_farm_index=8}: subspace_farmer::single_disk_farm::plotting: Replotting complete
2023-11-10T12:45:04.797730Z INFO single_disk_farm{disk_farm_index=8}: subspace_farmer::single_disk_farm::plotting: Replotting sector (0.00% complete) sector_index=0
look at just now, i am trying my best to tell you mine is in a loop and will not stop
It is not in a loop and it will stop. Provide all logs since node started, not just two lines selectively. You’re providing extremely limited amount of logs that are perfectly in line with expectations.
i am sorry i have no log files to post, only what i see scrolling. i will just wait it out since you say all is ok and working as it should.
Is that 2 log lines in a row? If it is, I agree this is really a problem to this plot. How can the log say replot is completed; and then in a few ms, it says it replot from sector 0 again???
Is that what you’re trying to say? Is it still happening now at your farmer’s log?
How about other plot?
yes , you can see from timestamp,and seen this over and over for 5 days now. same one completes, and then restarts replotting
and i have 170G and 325G plots, all doing same thing
i am still getting rewards during the replotting, so i will just wait see if ever does stop.
I do not have this issue, but I think I figure out why.
I’m farming with 1 plot only, my CPU is pretty good, a 10900, can make a sector in 4 minutes. Your CPU can be more powerful than me, but the most is 2x, i.e. it can make a sector in 2-3 minutes.
Now let’s say you have each plot 400GB, but I can see you put in your farmer at least 18 disks/plots.
So even it can create 1 sector in every 2 minutes. In total, it has 7200 sectors to fully plot for all your 18 400GB plots. Which means the time difference between sector 0 to sector 399 of any plot is approx 14400 minutes (240 hours). So, no wonder when it finishes sector 399, it immediately see sector 0 has already been expired.
The root cause is you put too much space in a single farm on early stage. I assume if you put more space at later stage, the time to replot will be longer, much longer.
so your recommending me to break down into multiple farmers on same pc. so 4 framers instead of 1 and have 8 plots per farmer?
There is no issue there, it works exactly as it should, already explained what is happening here:
Exactly.
Yes and no, situation will improve/get easier as network gets bigger. Also sectors start replotting a bit ahead of time, so in most cases you’ll farm with the whole plot anyway.
It will make no difference (at least no positive difference) as long as it is on the same machine.
thanks for the info, i will just let it ride out, did not want to break down to multi farmers anyway, i did stop all plotting on this 1 pc, and it will be a replotter/farmer only.
For best efficiency and least (unnecessary) resource usage you should run a single farmer with a single farm per physical disk. Any deviation from this may work, but will be suboptimal.
and that is what i am doing, got 24x200g ssds and 56x400g, and all 1 plot per ssd