While I was looking at the safenode-manager add help to see how to add the discord user I saw a very interesting thing
If the option (owner) is not used then the node will assign its own username …
To me this means that it acts as if I added a discord name and thus be returning nano to maidsafe for tallying.
This is not desired behaviour and you may find this is the source of the problem of why no discord username and still returns nanos to maidsafe
I am trying out safenode manager for the first time.
(minor) bug: After safeup update (to latest, e.g. safenode 0.106.5) and safeup node-manager (installing 0.8.0) and doing safenode-manager upgrade with no nodes existing I got this:
Retrieving latest version of safenode...
Latest version is 0.106.5
Downloading safenode version 0.106.5...
Download completed: /home/xxxx/.local/share/safe/node/downloads/safenode
Refreshing the node registry...
The application panicked (crashed).
Message: index out of bounds: the len is 0 but the index is 0
Location: /home/runner/work/safe_network/safe_network/sn_node_manager/src/cmd/node.rs:421
Backtrace omitted. Run with RUST_BACKTRACE=1 environment variable to display it.
Run with RUST_BACKTRACE=full to include source snippets.
Hey, thanks for raising the point.
I think I would need @qi_ma to clarify here on what the behaviour is if the owner is set by the node. I’m not necessarily sure there’s a big impact here, and we may just need to change the documentation to be clear that the owner assigned by the node is just an arbitrary thing that has no relation to Discord. Or just take the statement completely out of the node manager documentation.
I can’t remember the exact versions off the top of my head, but there was a breaking change between some recent versions of the node manager, which is probably what’s happened here. The node manager can’t deserialize your old registry file. I’m now being careful to make sure that newer versions of node manager are backwards compatible with old registry files.
it has to be the same discord_name that reported to us, otherwise your forwarded payments won’t get tracked, i.e. you will lose your rewards.
Meanwhile, as it is an arbitrary thing you choose by yourself, it doesn’t need to your real discord_name. just need to report it as your discord_name to us
Thanks for the help.
So what if no owner is assigned?
Is this the correct assessment? If it is not, the documentation probably needs to be changed.
it will then use the default owner, which is us.
no mater the owner is set or not, the received storage payment will be forwarded back to us.
Right, thanks. I guess I’m not completely sure about whether there is actually any problem here. Can you clarify please @neo ? Is it the wording of the documentation that has given you the wrong impression here?
The point is this
And this
tl;dr
Without specifying a owner, the help tells us that the node will assign its own discord name
THUS we cannot earn nanos by not giving a owner to the node (or node-manager)
@JimCollinson keeps telling us that we can earn nanos by not giving a owner and be able to use the nanos for uploading. BUT the way its written now the nodes cannot earn nanos to keep by simply not giving a owner
However: This was the first time ever that I tried safenode-manager.
I started first 3 nodes with the TUI and then 3 nodes more. Of the second batch 2 are deemed as running and one as added. But according to Vdash, they are all connected and working normally.
EDIT: after adding even more nodes, all of them are shown correctly as running 
All good folks.
We are sorting it.
@chriso sorry for being a broken record, asked this 3 times now (not to you in particular) can we set how many logs to keep via node-manager if not I will sort it out myself but I am sure it is a required feature.
To be clear, what the node manager actually says is this:
It doesn’t say that it will assign its own Discord username. However, I will add some extra text just to clarify what is going on.
I think the discrepancy between the balance in the node manager versus the earnings in the logs is a completely separate issue. The node manager is reading the balance directly from the wallet.
Sorry, I don’t think I had seen any requests for this.
I think there is an issue here in that log retention can be specified in terms of different things, like the size in MB, or number of files, etc. It could potentially get a little complicated if we wanted it to correspond exactly with everything that the logging framework offers.
What is the problem you have at the moment?
Not necessarily a current problem but I can see it becoming one with longer running networks. In the past I had node logs take up way more space than chunks.
We can currently (or were able to) set how many logs we keep when starting the node.
Not at my pc but if I remember correctly we could keep only current or however many archived logs as we choose.
I am hoping only to access that what is already implemented, via manager.
Ah ok, I see there are some arguments on the node related to this. Yeah, we can expose the same on the manager, no problem.
Mucho Gracias Sir.
The baseline is that my nodes without owner set still do not keep their earnings, they are being sent back to maidsafe. Everything else was explaining and discovering the source of the problem. The rest is said my qi_ma
Since owner is for discord name then that wording does in fact translate into discord name being for maidsafe. That is the logical outcome of the the help wording and qi_ma response.
This is the real issue at hand, not the wording of help or anything else. The node operator has no way to keep their earnings in this beta testing when @JimCollinson said we could by not specifying owner. And it seems this is being fixed, so I am just clarifying my previous words here.
This PR was merged:
In the next release of the node manager, by default, system-wide services will not start when the OS is rebooted, and for user-mode services, they will not start upon a new login. You need to use --auto-restart to opt in to that.
On Windows, WinSW will be configured automatically.