I just open up another tab in terminal and run it in there as normal. And use other tabs to do any cli I want to. I actually use another tab for all my node manager stuff, including doing a balance command to see my real earnings as vdash is rarely close
Any chance you can link me to the cargo.toml? My understanding right now is eg, you’re pulling in safenode and want to know the compatible networks for that bin? (or same for client?)
If that’s the case, simplest thing would be to expose an RPC call to get that info, I think.
you can install tmux and split the screen into multiple windows.
2 Likes
I can post it here later (it isn’t on github yet) but I shall clarify until I can do that: I’m not pulling in safenode (depending on what you mean there). I’m writing a client so I use various client side crates sn_client, sn_transfers etc which are either things I’m using directly (like sn_client, sn_registers), or things that need to be added to my Cargo.toml for the build to succeed.
I guess one way to do this might be to check-out sn_client at the relevant version in safe client app built for a testnet, and then grep its Cargo.lock for the modules I have in my Cargo.toml.
I’m not sure that’s adequate long term, but maybe I should just do that?
Aha, yeh that’s the ticket, the protocol is defined at build time inside of sn_client, so it’ll depend on the build vars and then the underlying networking version (as decided by sn_client version).
So probably the client shoudl have a protocol API to be able to surface that easily to you?
1 Like
Maybe, I’m not sure if that covers everything. Essentially you have the network protocol one side and the APIs the other. The APIs need to be in step with each other (and the app using them), and the APIs and protocol need to be in step with each other.
Thought I’d give this another go on my new laptop and Ubuntu 24 but the same result. cc @chriso
safeup client --version 0.91.2
safeup node --version 0.106.2
**************************************
* *
* Installing safe *
* *
**************************************
Installing safe for x86_64-unknown-linux-musl at /home/mrh/.local/bin...
Installing safe version 0.91.2...
⠐ [00:00:19] [#####################################>--] 11.21 MiB/11.86 MiB (1s) thread 'main' panicked at /home/runner/.cargo/registry/src/index.crates.io-6f17d22bba15001f/sn-releases-0.2.0/src/lib.rs:193:56:
called `Result::unwrap()` on an `Err` value: reqwest::Error { kind: Body, source: hyper::Error(Body, Kind(UnexpectedEof)) }
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace
**************************************
* *
* Installing safenode *
* *
**************************************
Installing safenode for x86_64-unknown-linux-musl at /home/mrh/.local/bin...
Installing safenode version 0.106.2...
⠂ [00:00:15] [#####################################>--] 8.94 MiB/9.56 MiB (1s) thread 'main' panicked at /home/runner/.cargo/registry/src/index.crates.io-6f17d22bba15001f/sn-releases-0.2.0/src/lib.rs:193:56:
called `Result::unwrap()` on an `Err` value: reqwest::Error { kind: Body, source: hyper::Error(Body, Kind(UnexpectedEof)) }
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace
Also, @joshuef will there be builds for Rasperry Pi or should we get ready to build ourselves for those and similar platforms (for beta)? Would be good to have a definitive topic on participation to say how people should prepare.
5 Likes
just thought i’d mention one of my nodes might now have become pritty pricy again …
[2024-05-17T15:27:41.347579Z DEBUG sn_node::quote] Created payment quote for NetworkAddress::ChunkAddress(611de0 - a01696ae1af1a6282389f956b2525afc4c4fcb3261f665bc4adebd837d46845f): PaymentQuote { content: 611de0(01100001)…, cost: NanoTokens(11490914232650), timestamp: SystemTime { tv_sec: 1715959661, tv_nsec: 347525802 }, quoting_metrics: QuotingMetrics { close_records_stored: 1743, max_records: 2048, received_payment_count: 28, live_time: 700649 }, owner: “user” }
[2024-05-17T15:27:41.627027Z DEBUG sn_node::quote] Verifying payment quote for NetworkAddress::ChunkAddress(611de0 - a01696ae1af1a6282389f956b2525afc4c4fcb3261f665bc4adebd837d46845f): PaymentQuote { content: 611de0(01100001)…, cost: NanoTokens(11490914232650), timestamp: SystemTime { tv_sec: 1715959661, tv_nsec: 347525800 }, quoting_metrics: QuotingMetrics { close_records_stored: 1743, max_records: 2048, received_payment_count: 28, live_time: 700649 }, owner: “user” }
11490914232650 nanos are … 11490 full token
4 Likes
Anyone running more than ten nodes on a pi4 8gb ram or similar? If so how is it running?
1 Like
Yes, I think you are right. I’ve just checked the log of my cost_checker script and there are some crazy valuations there. The worst I’ve seen is 112 tokens per the 4 chunks but your’s seems to top the lot. But it’s not all by any means. Most are still in the range of a couple of dozen SNT.
I think this happens when nodes start leaving the network. I’ve had to shutdown a few and killed a couple I had up for testing things. I think then there are ‘gaps’ in the coverage that aren’t being filled because not enough new nodes are joining. We can’t control where they join in the hashspace so we can be unlucky and they join where there is adequate coverage. Maybe there is a mechanism where things even out over time but my gut feel is that a gently shrinking network is not healthy from a cost point of view even if the fill level is still only around 60%.
5 Likes
Yes! I was running 40 on a RPi4 8GB (max I can realistically run at home with my 80/20 ADSL). No problem. I’ve had to reduce it to 20 after a suspicion that ‘the internet’ was being slow from some parties here who shall remain nameless. But now it’s the weekend there is less traffic and I’m going to gradually start some up again.
With 20 the CPU is still around 40% idle and it was about 25% idle with 40. I think it could have handled 50 or even 60 but I’ll never know from here. Not sure it would have been suitable for other tasks like media serving, browsing or being a PiHole or something at that level though.
I can run 10 no problem about 25% cpu but as soon as I up it to 20 my cpu goes through the roof. What’s the longest you had the 40 running?
1 Like
Sssssssshhhhhhhhuu… is coming for you 
4 Likes
hoping he will show me how to do it properly lol
4 Likes
Ran 10 on a rpi 3 with 1gb - just reduced node count because my Internet speed is crap…
1 Like
I have dropped from 20 to 15 going to see if it makes a difference
8 days in total and I stopped 20 of them 2 days ago to the hour.
I don’t know why yours are being murdered! Do you start them up with a bit of an interval? I find 2 mins is enough so --interval 120000.
I use Neiks script to start them. I have 30 running on a dell mini pc with similar spec and no problems runs about 50% cpu. The pi just keeps creeping up and up
Sorry not neik @aatonnomicc i forgot about your transition 
3 Likes
You better not miss gender me or get my pronouns wrong or I’ll set the @Southside on you
4 Likes