-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/41438/
-----------------------------------------------------------

(Updated Dec. 16, 2015, 8:27 a.m.)


Review request for mesos and Bernd Mathiske.


Bugs: MESOS-4130
    https://issues.apache.org/jira/browse/MESOS-4130


Repository: mesos


Description
-------

Added documentation on using network proxy for mesos fetcher


Diffs
-----

  docs/fetcher.md ec7372293491ed9ee4e20dda3c2def72a77a3f84 

Diff: https://reviews.apache.org/r/41438/diff/


Testing (updated)
-------

- Set `https_proxy=http://localhost:3128` (a local squid server) in 
`/etc/default/mesos-slave`, restart mesos-slave, launch a new marathon task 
with a `https://` uri, and through the following sysdig command I can see there 
is data exchanging between mesos-fetcher process and port 3128.

```
sudo sysdig -A -c echo_fds proc.name=mesos-fetcher and fd.port=3128
```

- Stop the local squid server, try to restart the marathon task, the task would 
fail repeatly, from slave logs there are error messages that fetcher failed to 
fetch the uri.
```
Dec 16 15:16:35 lin-E400 mesos-slave[24247]: E1216 15:16:35.678032 24283 
slave.cpp:3342] Container '45c14132-c56a-4cff-a6b5-f57ba2670643' for executor 
'testapp_web.f0ef72d2-a3c4-11e5-af60-56847afe9799' of framework 'b52179fd-8968-4
bf8-baf0-dddc8a38c903-0000' failed to start: Failed to fetch all URIs for 
container '45c14132-c56a-4cff-a6b5-f57ba2670643' with exit status: 2
```

- Restart the squid server, the task would start without a problem.


Thanks,

Shuai Lin

Reply via email to