I would suggest a wiki in a category that is only readable by TL1 and above. That way it is not scanned by robots incl search engines.
4 Likes
Still seems risky, maybe an unnecessary risk with a vital project resource.
Say the forum gets taken down by the copy lobby … 
I’d Safe it for the network.
2 Likes
Yes, I don’t think it’s smart to involve Maidsafe in something that, almost certainly, will contain copyrighted material.
If something is raised, I think it should go outside this forum. Telegram channel, for example.
3 Likes
Dare we ask when we could host such things somewhere other than the forum or existing internet. Just this one “when” as a special occurrence 
Being a total noob here, but I have not been successful in connecting a node.
I forwarded a port range of 12000-12099 to my local Raspberry Pi 2b, and updated client and node every time before trying.
Then I launched a node with
safenode --port 12000
Which starts, but never connects to anything.
When I try now, with the latest version, I see the following error:
safenode --port 12000
Logging to directory: "/home/rpi/.local/share/safe/node/12D3KooWF4bFwSwkwgYhvTsvwG8EeFmo6DS1CfW4KG8ntqPw46W1/logs"
Trying to fetch the bootstrap peers from https://sn-testnet.s3.eu-west-2.amazonaws.com/network-contacts
Node started
PeerId is 12D3KooWF4bFwSwkwgYhvTsvwG8EeFmo6DS1CfW4KG8ntqPw46W1
You can check your reward balance by running:
`safe wallet balance --peer-id=12D3KooWF4bFwSwkwgYhvTsvwG8EeFmo6DS1CfW4KG8ntqPw46W1`
The application panicked (crashed).
Message: cannot sample empty range
Location: /cargo/registry/src/index.crates.io-6f17d22bba15001f/rand-0.8.5/src/rng.rs:134
Backtrace omitted. Run with RUST_BACKTRACE=1 environment variable to display it.
Run with RUST_BACKTRACE=full to include source snippets.
The application panicked (crashed).
Message: cannot sample empty range
Location: /cargo/registry/src/index.crates.io-6f17d22bba15001f/rand-0.8.5/src/rng.rs:134
Log here:
safenode.zip (9.9 KB)
I would like to see folks not sharing copyrighted material. I think that at this point of testing there is a chance that it will work against our goals. It may be a small chance, but nevertheless.
The benefit of music and movies is of course that many of us have them at our computers and they are easy to get for that reason. It’s also easy to validate if they work after downloading, or not. Could we find some free to use files that are easy to validate? Linux.iso’s are not that easy.
5 Likes
There’s a topic listing public datasets.
Figuring out ways to upload and share each of those would also be good preparation for launch, and great for marketing.
Much better all round if the community are doing a public service as the network begins to attract interest.
A few scripts to copy of data from that list and share its location on SN would be a good place to start.
4 Likes
Upload your current file list as a text file to the testnet and list that file. Every time you upload a few files upload a new copy with all the files you uploaded and post that new file list file.
That way its kept on Safe and you’re just posting here the details to download that list file. (and each update as you upload a new one) Is append records implemented, if so then just keep appending to an initial record.
This all seems very convoluted.
The system where Maidsafe uploaded a bunch of data a while back and added it to the op seems a simple solution.
Anyone who wants to test downloads grabs from there, those who want to share additional non copyright stuff posts a link.
If you want to upload terabytes of copyright material to test downloads keep it for your own use.
Really seems we are making a mountain out of a molehill.
7 Likes
Good points!
That way we also get to test the oldest data on the network, which is great.
4 Likes
I think that I asked this before, I think the answer was not a problem.
If you have a dynamic IP and it changes midway through running a node, issue?
1 Like
I don’t remember you asking that before and I don’t remember what the answer was. But I’ve just tested that because it is of interest to me as well and I didn’t like my IP because it wasn’t all prime numbers.
My nodes seem fine with it because I’m getting the ‘get_closest returned’ messages again.
3 Likes
love it 
1 Like
Thanks for everyone involved in this testnet. It’s been a real good one and very helpful!
I’ll be bringing the maidsafe nodes down soon.
16 Likes
Updating the node to 0.96.3 fixed the issue, it does not crash anymore.
Still can’t connect to anything though.
1 Like
To my mind the point at which the nature of files uploaded for testing purposes could be said to be problematic might be at two points:-
- when there is a public beta which will hopefully attract more attention than a small group of enthusiasts.
- when a testnet persists for a longer time than a couple of weeks. At the moment they are torn down after they have been tested to destruction or there is a fresh one planned. When the data is expected to persist indefinitely that could be judged to be a different matter.
2 Likes
Now that this testnet is winding down I’ve taken the graphs generated from @Josh 's NTracking system. NTracking
Archive.zip (7.1 MB)
Here are some highlights
TCP Connections
Something happened on 21/10 lasting for more than a day. More than one node had a big drop in tcp connections.
Bubble Rewards
There’s the big boy. But there were a couple of other nodes that were doing well.
Rewards Balance
My super earner node with 160 Toy Town Tokens dwarfed all other nodes apart from one that had 43.
You have to switch off the top 3 to be able to see the growth of the others.
All we can say for sure is that rewards over time for a node and rewards over record count are unpredictable. Something that is apparently doing poorly can get a rocket up it. This may all be due to the smallness of the network.
Record Count
The record count was actually quite even in the end.
Memory Usage
There is still some kind of memory leak but it’s a lot less than before. Those big peaks for a few nodes tie in with when there were a lot of GETs. The values did decrease but not completely.
I started 40 nodes on a AWS T4g.medium Instance with 4GB RAM. I was kicking myself for not being bolder and starting 50 on it when it was clear the memory leak issue was a lot less. However, but the end there was only 500MB available so I’d have had some nodes killed.
13 Likes
I am clearly way to conservative.
I thought I was going to live on the edge and run 60 on 8GB of RAM. maybe I’ll do 70.
2 Likes
In preparation for the next testnet, question for the community from a linux noob:
Any suggestion on how to have nodes and client store their data in a separate volume?
Problem: my cloud instance has avery small root volume capacity, while the bulk of the storage is in a separate block, currently mounted to /volume1/ directory.
On that volume I also store all files to upload for test.
All node and client data is stored in the root volume, under ~/.local/share/safe, and that fills up my root volume, so I can’t run more than just a few nodes.
I suppose I could mount the volume to the local/share/safe location, but then what happens during clean up, when I remove that folder?
2 Likes
A few options here.
One is to run the node with the --root-dir option to the location of the volume, eg
safenode --root-dir /media/mydevice/safenodes
Another is to mount the volume to the default location for node storage
sudo mount /dev/sdb1 /home/user/.local/share/safe/node
Another option is to point the default node storage directory to the drive using a simlink
ln -s /media/mydevice/safenodes /home/user/.local/share/safe/node
I reckon the first option is the simplest
7 Likes