CDN trouble in 'fetch v8'?

[Dan Kegel]

I'm debugging a v8 build script.  All was well the first couple days, but now

'fetch v8' is failing with 

 pip._vendor.requests.exceptions.ConnectionError: ('Connection aborted.', error(65, 'No route to host'))

or

pip._vendor.requests.exceptions.ConnectionError: ('Connection aborted.', error(101, 'Network is unreachable'))

depending on which OS I run it on.

netstat -tp | grep python shows the host it's trying to connect to is

170.216.58.216.in-addr.arpa domain name pointer sea15s02-in-f10.1e100.net.

Is anyone else having problems?

Also, may I just say that 1e100.net is a very cute domain name?  :-)

[Ben Noordhuis]

FWIW, it worked for me just now. Maybe you can use something like
`ssh -D` + tsocks to route the traffic over another host if the
problem persists on your local machine?

[Dan Kegel]

On Fri, Jun 26, 2015 at 2:59 AM, Ben Noordhuis <in...@bnoordhuis.nl> wrote:
>> 'fetch v8' is failing with
>> pip._vendor.requests.exceptions.ConnectionError: ('Connection aborted.',
>> error(65, 'No route to host'))
>

> FWIW, it worked for me just now. Maybe you can use something like
> `ssh -D` + tsocks to route the traffic over another host if the
> problem persists on your local machine?

That's an idea.

And I should definitely do local caching.

Still failing today, btw. I'll try running it on some other (distant)
hosts and cache the result.
- Dan

[Dan Kegel]

Doing it on a small DigitalOcean instance succeeded, at least after I
turned swap on :-)