Hah, and now I see that my router has dropped all the 20 port forwardings set by upnp! And all my nodes seem to be rolling nicely in TUI and Vdash, but only one of them has earnings.
I checked that my laptop does not have any settings that would put it to sleep after a period of time etc. So probably it’s my router doing it’s own thing.
If things like this happens to many others, I can see troubles ahead…
I wonder if nodes could detect if the port forwarding fails, and react to that by changing into home network -setting?
Prob better to have a uPnP keep alive so that if the node hicups for enough time for the router to forget the uPnP then the keep alive ensures its not off for very long when the hicup finishes
Hicup could be as simple as drive powered down and node s/w is waiting for it to spin up again. And this can be over 10 seconds in more extreme case or NAS that spun down
Nope It’d have to be in the node s/w. Already mentioned it to the devs so we just have to wait for that. If the router allows you to extend the uPnP time then maybe you could change that but I am clueless on that since I’ve never dealt with that before
Yeah, tried to look for something like that, but didn’t find. Also it dropped them after just a few hours, when it previously has kept them much longer.
Then again the nodes seem to be working very well without upnp. But if I have understood right, node utilizing relays are bit of a burden to a truly public nodes. I hope people don’t need to make a choice of serving themselves or greater good.
I have a choice of Open this device completely for internet sharing via IPv4 (exposed host) that I could apply for one device from my router. I could do that for one laptop that has nothing important in it. Or would it be a security risk wider than the machine itself? Could someone gain access to my home network via that machine?
I’m getting the following on the above command and I’m trying to figure out what it means. Because it fails to add the RPC port of choice, it gives a random one (36483) and I have to point the router to it?
I believe you’re missing -- before rpc-port
But in any case, unless you collect metrics, I don’t think opening those port is required for the node to work properly
I’m struggling with the launchpad for Windows 10 22H2 build 19045.4412 in my case.
I downloaded winsw.exe into system32. But after allocating resources and setting Discord username when starting nodes (launchpad under admin privileges) it:
creates launchpad folder at “C:\Users\Username\AppData\Roaming\safe\launchpad” with app_data.json file and launchpad.log file (attached) launchpad.zip (1.9 KB)
downloads “safenode-0.106.5-x86_64” into ProgramData folder
creates “data” and “logs” folders for each node there
puts “safenode.exe” into each nodes’ folders
But it doesn’t create any wallet files and safenode.pid thus nodes don’t even try to start it seems (no safenode.log for any node).
P.S. Add my Discord username 0belius to the list. I have 7 nodes on my Macbook as well from yesterday.
P.P.S. It would be great if we could set a different hard drive on our choice in launchpad thus I could allocate more than 15 Gb on my windows PC
Hi @Zoki I think that the launcher will start nodes based on the size allocated. If you want to reduce the number of nodes, I recommend using the safenode-manager to stop and start the nodes as needed.
how’s every one getting on cpu wise I had a period last night where where things went a bit mental. had all systems in the high 90% things seem to now have settled down to more normal cpu levels of 50% to 70% CPU.
quick question @joshuef how does a reboot of a machine affect the nodes on it if they are run as a service ? do they all start from where they left off or do they restart from zero?
asking as I had one machine this morning that was jammed on at 100%.
running 50 nodes on a 4 vcp 6gb ram 200gb HD on 250/250 internet connection.
logging in took ages via ssh and I rebooted it and now it is using less cpu than the rest and comparable with when all the nodes were first started at 48% CPU.
I had a period of several hours at the left of the graph of high cpu followed by a another period this morning.
the red line along the top was the vps that I restarted after never coming back down. which after a reboot has come back in line with the rest of its pall’s.
interested to know what other people are seeing and possibly what the team are seeing on there DO vps’s if any one has time?
can see around the same time period someone was smashing around about 700 nodes.
from 22:00 till about 03:30 my time. which coincides with my heavy CPU usage patch.
Yeah, I also have 8GB of RAM which is a bit silly and overkill.
Not all cores are equal though but yes I could probably do 40. If they can cruise through the ups and downs of network temperament without displacing my zen I will give them some more responsibility
Am happy to max my machine out and fry eggs on it.
