I’ve disabled UPNP on all of my harvesters in the chia config.
Also checked this on all my harvesters…
$ dpkg -l linux-igd
dpkg-query: no packages found matching linux-igd
UPNP is not enabled on my router, well, only on the VLAN that has my gaming consoles. Yes I hate UPNP so much that I have a VLAN for just my consoles so I can enable UPNP.
Should I disable the UPNP in the farmers config?
The only reason I didn’t disabled UPNP on my ubuntu harvesters is because there was nothing mentioned in the chia guide like there was for windows.
I thought disabling upnp needs only been done, when you have multiple nodes. Afaik, if a Harvester is launched by CLI, its just a harvester not a full-node. Therefore it does not need upnp to be disabled specifically. If you launch the GUI, it actually is a full-node and therefore should have upnp disabled (if there are more than one full-nodes).
I thought this to be true also. The upnp setting is under the full node section of the config. I’m only running one full node and the rest are just running harvester.
I do this option #1, except I share my plotter’s plot directory over LAN. That way I see the new plots show up on farmer as they are built, and I do no transfers over network, causing slowdowns. When the drive filles, I umount and take to the farmer and mount.
Also I don’t run any chia daemons like a harvester, node, at all, on the plotter. Its just running chia plots create.
@mxfh - I find the same thing , no difference over GB LAN. 1G LAN is nearly the same speed as a spinning SATA disk anyway, it shouldn’t be any issue if you have your network setup and running optimally. The only thing maybe I would see is if 3-4 plots are writing at once, which would slow the disk down, not the LAN. I won’t do plot transfers over LAN, too slow, slows down the farmer a bit I think too.
FYI: I was fairly concerned about switching over to a multiple harvester model after reading this and a couple other posts. I had some timings starting to creep up so finally switched over to multiple harvesters yesterday. I followed all the tips I found here and verified all the logs several times. I was still a bit worried but then I hit a Proof this morning with the new setup!! (about 12-hours later) So I would say if done correctly and you follow all the tips on chiaforums, you should be good…
For those interested, here’s how the harvester works:
The harvester receives a new signage point from the farmer, this happens at the start of each slot.
The harvester applies the plot filter for each of the plots, to select the proportion which are eligible for this signage point and challenge.
The harvester gets the qualities for each plot. This is approximately 7 reads per plot which qualifies. Note that each plot may have 0, 1, 2, etc qualities for that challenge: but on average it will have 1.
Checks the required_iters for each quality and the given signage point, to see which are eligible for inclusion (required_iters < sp_interval_iters).
Looks up the full proof of space in the plot for each quality, approximately 64 reads per quality
Think you got my luck I “missed” two blocks according the average expected reward. I completely moved to one farmer with separate harvesters now.
I find it kind of strange the farmer (gui wallet) is not reporting anything about my harvesters except that they are connected in the advanced farm window.
I set my harvesters to INFO logging and now I do see these kind of lines in my logging:
2021-06-07T14:06:10.651 harvester chia.harvester.harvester: INFO 2 plots were eligible for farming 8basf39692… Found 0 proofs. Time: 1.38876 s. Total 1585 plots
I hope that’s enough proof the system is working as it should, but if someone can confirm this, that would be nice. Until I find a new block one day I still don’t feel 100% comfortable about this setup yet.
chiaharvestergraph is really a useful monitoring tool. Still I am not 100% sure my farm is working properly now. (see a few messages above) Will keep you guys posted when I find a block with this new multiple harvester setup.
Set the log level on your farmer to DEBUG and you should see messages coming from you harvesters in the farmer log. You can just search for the IP addresses of the harvesters.
Well said. Hope there is some progress. We on 10k plots on the LAN share method, but constantly have “harvester not running” issues, or too slow to respond. The multiple harvester method works, in theory, but not having high level status is stressful.
chia farm summary ought to work for multi-farmer setups.
In the end I found a block as well on one of the external harvesters. Think I had just bad luck, but will keep you updated here if I still have an issue.
The club of people with > 1PiB plots and no rewards for > 1 month is too much of a coincidence. There must be some form of explanation for this other than “bad luck”.
Anyone with > 1PiB of plots who win regularly? If so, what version of chia are you running? How many computers are you using etc.?
I try to put everything on one farm with network connection to the other units on server / nas.
If I manage to get times of less than 30 seconds, I wait a few days and see if the winnings arrive. If they do, a problem is evident
I “missed” two blocks according the average expected reward. I completely moved to one farmer with separate harvesters now.