When trying to connect to tensorboard logs on a remote server and entering the address http://localhost:16006/ in chrome and firefox I get this message in the command line "channel 3: open failed: connect failed: Connection refused" multiple times and this message on the browser, "The Connection was Reset".
I ssh into the server like this: ssh -L 16006:127.0.0.1:6006 username#machine and then go one level up the log folder and write: tensorboard --logdir logs --port 16006
I tired:
tensorboard --logdir logs --port 16006 --bind-all
and also
tensorboard --logdir logs --host localhost, also
tensorboard --logdir logs --host 127.0.0.1
None of the above has wroked. I tried running the line above from another environment which didn't help. I went to the office and tried connecting to the logs from the server machine directly and it worked.
It used to work before when accessing remotely. Do you know what the problem is? any hint would be immensely appreciated.
I am having the same problem, but I think you should designate --port 6006 since it looks like you are forwarding remote port 6006 to the local port 16006.
Even so, since 6006 is the default port, the other commands should have worked, but you would have to go to http://127.0.0.1:16006 on your local machine, rather than the link it provides.
Some more in depth explanations can be found here how to run tensorboard on a remote server and how to see tensorboard over ssh
Even following this advice though I am still getting a 'channel 3: open failed: connect failed: Connection failed' error
Related
I'm trying to run a storybook server on AWS Cloud9 but the URL it gives doesn't load anything.
I'm starting the server with
start-storybook -h $HOST -p $PORT --ci
This runs through without error and gives me a "server started" message with a URL. But that URL doesn't connect to anything.
I do notice that the URL isn't secure, and I can imagine AWS having issue with that. There is an --https option on the start-storybook command, but it requires SSL information that I don't know how to source.
Anyone know how this I can get this working?
C9 only opens port 8080, 8081, 8082. So your server should be listening on one of three. Try:
start-storybook -p 8080 s public
I'm trying to set up an SSH tunnel to access my server (currently an ubuntu 16.04 VM on Azure) to set up safe access to my django applications running on it.
I was able to imitate the production environment with Apache WSGI and it works pretty good but since I'm trying to develop the application I don't want to make it available to broader public right now - but to make it visible only for a bunch of people.
To the point: when I set up the ssh tunnel using putty on Windows 10 (8000 to localhost:8000) and I run http://localhost:8000/ I get the folowing error:
"Not Found HTTP Error 404. The requested resource is not found.".
How can I make it work? I run the server using manage.py runserver 0:8000.
I found somewhere that the error may be due to the fact that the application does not have access to ssh files, but I don't know whether that's the point here (or how to change it).
Regards,
Dominik
After hours of trying I was able to solve the problem.
First of all, I made sure putty connects to the server and creates the desired tunnel. To do that I right-clicked on the putty window (title bar) and clicked event log. I checked the log and found the following error:
Local port 8000 forwarding to localhost:8000 failed: Network error:
Permission denied
I was able to solve it by choosing other local port (9000 instead of 8000 in my instance).
Second of all, I edited the sshd_config file: sudo vi etc/ssh/sshd_config
and added these three lines:
AllowAgentForwarding yes
AllowTcpForwarding yes
GatewayPorts yes
I saved the file and restarted the ssh service:
sudo service ssh stop
sudo service ssh start
Now when I visit localhost:9000 everything works just fine.
Im pulling my hair out with this one. I've managed to get a site running on Digital Ocean through Forge and also SSH into the server so I know that the SSH keys are setup correctly.
But when trying to connect via Navicat, I keep getting the error "Unable to load key". Does anyone know what this is in reference to or what the problem is.
regards
There is an answer from 2015 on Navicat forum, indicating there could be issue with ssh key support.
If you can connect via ssh, do an ssh tunnel to your server manually:
ssh -v servername -L 3306:127.0.0.1:3306 -N
Afterwards, you connect to remote mysql as it was on your localhost (you should not have one running on 3306, otherwise map to another local port)
I am experiencing issues with the latest bosh-lite virtual box machine. See here.
I have just downloaded the Vagrantfile and done a
vagrant up
Then a:
cf login -u admin -a 192.168.50.4 -p admin
But it give me a:
API endpoint: 192.168.50.4
FAILED
connection refused
Can anyone please help?
Get address of haproxy by logging to it (bosh ssh, then ifconfig). Use address of haproxy as api endpoint.
I just begun using Flask and tried to run the hello_world example (hello.py)
The point is that I run the server through ssh on a remote machine and I want to browse it from my local machine so I used
app.run(host='0.0.0.0'),
However when I browse on chrome I put let's say http: //RemoteMachineIP:5000/ I got:
Oops! Google Chrome could not connect to RemoteMachineIP:5000
The remote machine has multiple ethernet IP addresses so I tried them (as http: //RemoteMachineIP:5000/) and I am always getting the same error.
I have also tried using app.run(host='LocalMachineIP') in hello.py but I got this error Flask [Errno 99] Cannot assign requested address.
Am I missing something ?
If you can run the server but the browser won't see it, it is a firewall problem.
If you can't assign the ip port, probably there is a PID already using the port, so you should try
lsof -i :5000
Then
kill -9 [PID From the lsof]