My dog is enjoying his new found fame too!
Installed and connected on Ubuntu 20.04 without any issue. Downloaded the test file just fine.
Can you try to get some other non existent file in the same way? We shoudl never see a stack overflow (well we can, but itās a bug).
I do not understand you. Glitch is happens with put
.
I will try file #3 next.
932210 bytes .png
file fails too.
I have changed account and 2nd acc fails too.
It may be Windows specific problem by the way.
Will try to get log.
Maybe it can help.
Last lines were:
[rustls::client::hs] TRACE 2021-04-13T19:36:33.111110500 [C:\Users\runneradmin\.cargo\registry\src\github.com-1ecc6299db9ec823\rustls-0.19.0\src\client\hs.rs:398] Starting early data traffic
thread 'main' has overflowed its stack
Hmm, wonder if this is a tls bug?
Is it possible to get build with debug information? .pdb
I can learn Rust and make build by myself, but it may take too long time.
Stack is not looking like looping, just too many calls.
Ubuntu cloud node is waiting to join.
Ubuntu 20.04 desktop working:
safe cat safe://hygoyeye9mq5jmipo3we79wohue8hjyw55e6f8xyk1hquwy9hsxnk9tdsme | display
safe cat safe://lazydog | display
safe cat safe://hygoyeye9mq5jmipo3we79wohue8hjyw55e6f8xyk1hquwy9hsxnk9tdsme > ~/safe-the-planet.png
$ safe auth create --test-coins
Passphrase:
Password:
Sending request to authd to create a Safe...
Safe was created successfully!
$ safe auth status
Sending request to authd to obtain a status report...
+------------------------------------------+-------+
| Safe Authenticator status | |
+------------------------------------------+-------+
| Authenticator daemon version | 0.5.1 |
+------------------------------------------+-------+
| Is there a Safe currently unlocked? | Yes |
+------------------------------------------+-------+
| Number of pending authorisation requests | 0 |
+------------------------------------------+-------+
| Number of notifications subscribers | 0 |
+------------------------------------------+-------+
QUESTION: I though I had to safe auth unlock
before the Safe was unlocked last time?
PROBLEM: opened a new terminal and repeatedly getting the following:
$ safe auth unlock --self-auth
Passphrase:
Password:
Sending action request to authd to unlock the Safe...
Safe unlocked successfully
Authorising CLI application...
Waiting for authorising response from authd...
Error: Application authorisation failed
Caused by:
AuthdError: AuthenticatorError: Failed to authorise application on the network: ClientError: Connection lost
Something new to download, hereās a new container with 11M of goodness to list and listen to:
safe cat "safe://hyryyry3bejyck7soabzf1esh6k1x5idx4zjyz4hkr438fkzm5pq6gxiezrnra"
And even better:
$ safe cat safe://hy8aynyc655n5qnhcb4dht6qhj9sifow6x19qj5tyapdd76epjot5mxadrh > yes-mp2-really.mp2
$ firefox yes-mp2-really.mp2
safe-the-planet and lazydog retreived here ( linux x64 )
and
~$ safe auth status
Sending request to authd to obtain a status reportā¦
Ā±-----------------------------------------Ā±------+
| Safe Authenticator status | |
Ā±-----------------------------------------Ā±------+
| Authenticator daemon version | 0.5.1 |
Ā±-----------------------------------------Ā±------+
| Is there a Safe currently unlocked? | Yes |
Ā±-----------------------------------------Ā±------+
| Number of pending authorisation requests | 0 |
Ā±-----------------------------------------Ā±------+
| Number of notifications subscribers | 0 |
Ā±-----------------------------------------Ā±------+
It looks like this thing is working again.
If you install rust, grab the repo and just do cargo build
it will be a debug build and you can connect a debugger to it. Itās a bit of work, letās see if others get it and we can try and recreate it in hose as well. Thereās no way some program has changed your stack size?
got mac setup up easily, node waiting to join. fetches well! nice onnnne
Trying to upload a 733 MB movie.
Sascha@sascha-Librem-15-v3:~$ time safe files put ~/Safe_put/Eraserhead.avi
Error: NetDataError: Failed to PUT Public Blob: QuicP2p(Connection(TimedOut))
real 8m22,145s
user 1m26,268s
sys 0m42,672s
EDIT: Tried again. No cigar.
@dirvine looks like I know what happened.
Not a loop. Just not enough stack.
I have patched safe.exe
: at loc 0x162
changed 0x10
to 0x20
and bug is gone.
Wild!
Yaāll are moving so fast I think you may actually be in the future!!
Try uploading that gif!
Hmmm, the windows stack issue is back perhaps (or some variation)??
Weād seen this something like this a while back, as by default windows stack was 1mb vs 8mb on linux/mac. we solved this by shimming the main
loop into specific sized threads to get around this defaultā¦
https://github.com/maidsafe/sn_node/blob/master/src/bin/sn_node.rs#L40-L47
I wonder whatās causing it this time
13mb file:
FilesContainer created at: "safe://hyryyryhs5rin4t5krr5yw1d89bhthf6u51wamh6i9n5fbwetck6f7i8fthnra"
+ DSCF2166.jpg safe://hygoygymu6dhdnyqe1ycyh8o4fnyz5rfyohsk9ewinjnu9qz5exebmq7jqo
35.09s user 4.19s system 33% cpu 1:57.16 total
and
time safe cat safe://hygoygymu6dhdnyqe1ycyh8o4fnyz5rfyohsk9ewinjnu9qz5exebmq7jqo > ~/x.jpg
16.06s user 7.64s system 43% cpu 53.952 total
Stack size can be tuned right inside PE header.
It will be more proper solution if you really need such large stack.
Once more: it is not in code, but in header.
Maybe Rust have compilation flags to change default value.
SizeOfStackReserve
here:
File put worked too
safe cat safe://hygoygycfm6iarj9j9otcrhzjhcxbuhprmn1fe3cktysahggg5u4yst1k9r | display