[General] data center machines unusable

[vmx]

What do you need?

I can't connect to most of the worker-gpu-* machines anymore. I was able to connect to worker-gpu-5, but that machine seems to have DNS/networking issues. The issue was confirmed by @cryptonemo.
=> They are not usable and I'm blocked as I can't do my work on those machines.

Why do you need it?

Please have a look and make sure that I can connect to those machines and that they don't have networking issues.

Who is the DRI?

@vmx and @cryptonemo

Team and command structure

FilCrypto

Estimated monthly cost

N/A

What else do we need to know?

That's all.

[vmx]

Update, I also cannot connect to miner-2 or worker-cpu-2-2. Though I can connect to worker-cpu-2-1. This seems to be a bigger data center issue.

[vmx]

I should also note that 24h ago things still worked as expected.

[vmx]

I still can't ssh to e.g. worker-gpu-6, but worker-gpu-5 doesn't have to seem networking issues anymore.

[ognots]

miner-2 should work now, I just rebooted it. it was wedged.
I was able to connect to the following machines and validate your user exists

• worker-cpu-2-1
• worker-cpu-2-2
• worker-gpu-6
• worker-gpu-5
  Can you paste your SSH config?

[vmx]

Thanks a lot!

• worker-cpu-2-1
• worker-cpu-2-2
• worker-gpu-6
• worker-gpu-5

I can now ssh to those above. Though I cannot ssh to:

• miner-2
• worker-gpu-3
• worker-gpu-4
• worker-gpu-7
• worker-gpu-8 (though that might not be in service any more, I haven't tried in a long time)

[vmx]

The networking still isn't good. E.g. on woker-gpu-6 I need several retries to pull from GitHub. The error is something like:

fatal: unable to access 'https://github.com/filecoin-project/rust-fil-proofs/': Failed to connect to github.com port 443: No route to host

[vmx]

worker-cpu-2-2 has the same networking issues.

[vmx]

Any news? The worker-gpu-6 still has those networking issues.