Hi, When I try to run github hosted regression tests with run.py, access to raw.github.com times out even though it seems to be using the specified our proxy server. # Running regressions in /scratch/modular-boost... # Creating regression scripts at /scratch/modular-boost/tools_regression_src... # Dowloading regression scripts from https://raw.github.com/boostorg/boost/develop/tools/regression/src... --- http://wwwproxy.sandia.gov:80 What's interesting is that I'm able to clone from github.com just fine, as you can see here. -bash-4.1$ git clone https://github.com/boostorg/boost Initialized empty Git repository in /scratch/modular-boost/boost/.git/ remote: Counting objects: 180556, done. remote: Compressing objects: 100% (68339/68339), done. remote: Total 180556 (delta 111934), reused 180501 (delta 111879) Receiving objects: 100% (180556/180556), 68.21 MiB | 10.23 MiB/s, done. Resolving deltas: 100% (111934/111934), done. But we're not able to clone from raw. -bash-4.1$ git clone https://raw.github.com/boostorg/boost Initialized empty Git repository in /scratch/modular-boost/boost/.git/ error: The requested URL returned error: 403 while accessing https://raw.github.com/boostorg/boost/info/refs fatal: HTTP request failed which means this clone (https://raw.github.com/boostorg/boost/develop/tools/regression/src) will fail. Is this a known issue and / or is there a workaround for this? This is on RHE Linux with git 1.8.5.1. -- Noel Belcourt
Does a regular wget
https://raw.github.com/boostorg/boost/develop/tools/regression/src/regressio...
Does it work with the proxy or without the proxy?
Note.. You can't use git tools with the raw URLs. They are meant to just
read the content of files directly. But you should be able to use web tools
(wget and browsers) with them.
On Wed, Jan 22, 2014 at 2:59 PM, Belcourt, Kenneth
Hi,
When I try to run github hosted regression tests with run.py, access to raw.github.com times out even though it seems to be using the specified our proxy server.
# Running regressions in /scratch/modular-boost... # Creating regression scripts at /scratch/modular-boost/tools_regression_src... # Dowloading regression scripts from https://raw.github.com/boostorg/boost/develop/tools/regression/src... --- http://wwwproxy.sandia.gov:80
What's interesting is that I'm able to clone from github.com just fine, as you can see here.
-bash-4.1$ git clone https://github.com/boostorg/boost Initialized empty Git repository in /scratch/modular-boost/boost/.git/ remote: Counting objects: 180556, done. remote: Compressing objects: 100% (68339/68339), done. remote: Total 180556 (delta 111934), reused 180501 (delta 111879) Receiving objects: 100% (180556/180556), 68.21 MiB | 10.23 MiB/s, done. Resolving deltas: 100% (111934/111934), done.
But we're not able to clone from raw.
-bash-4.1$ git clone https://raw.github.com/boostorg/boost Initialized empty Git repository in /scratch/modular-boost/boost/.git/ error: The requested URL returned error: 403 while accessing https://raw.github.com/boostorg/boost/info/refs fatal: HTTP request failed
which means this clone ( https://raw.github.com/boostorg/boost/develop/tools/regression/src) will fail. Is this a known issue and / or is there a workaround for this? This is on RHE Linux with git 1.8.5.1.
-- Noel Belcourt
_______________________________________________ Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost
-- -- Rene Rivera -- Grafik - Don't Assume Anything -- Robot Dreams - http://robot-dreams.net -- rrivera/acm.org (msn) - grafikrobot/aim,yahoo,skype,efnet,gmail
On Jan 22, 2014, at 2:09 PM, Rene Rivera wrote:
Does a regular wget https://raw.github.com/boostorg/boost/develop/tools/regression/src/regressio... Does it work with the proxy or without the proxy?
Unfortunately no, neither works.
Note.. You can't use git tools with the raw URLs. They are meant to just read the content of files directly. But you should be able to use web tools (wget and browsers) with them.
Okay, thanks for the info.
On Wed, Jan 22, 2014 at 2:59 PM, Belcourt, Kenneth
wrote: Hi,
When I try to run github hosted regression tests with run.py, access to raw.github.com times out even though it seems to be using the specified our proxy server.
# Running regressions in /scratch/modular-boost... # Creating regression scripts at /scratch/modular-boost/tools_regression_src... # Dowloading regression scripts from https://raw.github.com/boostorg/boost/develop/tools/regression/src... --- http://wwwproxy.sandia.gov:80
What's interesting is that I'm able to clone from github.com just fine, as you can see here.
-bash-4.1$ git clone https://github.com/boostorg/boost Initialized empty Git repository in /scratch/modular-boost/boost/.git/ remote: Counting objects: 180556, done. remote: Compressing objects: 100% (68339/68339), done. remote: Total 180556 (delta 111934), reused 180501 (delta 111879) Receiving objects: 100% (180556/180556), 68.21 MiB | 10.23 MiB/s, done. Resolving deltas: 100% (111934/111934), done.
But we're not able to clone from raw.
-bash-4.1$ git clone https://raw.github.com/boostorg/boost Initialized empty Git repository in /scratch/modular-boost/boost/.git/ error: The requested URL returned error: 403 while accessing https://raw.github.com/boostorg/boost/info/refs fatal: HTTP request failed
which means this clone ( https://raw.github.com/boostorg/boost/develop/tools/regression/src) will fail. Is this a known issue and / or is there a workaround for this? This is on RHE Linux with git 1.8.5.1.
-- Noel Belcourt
_______________________________________________ Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost
-- -- Rene Rivera -- Grafik - Don't Assume Anything -- Robot Dreams - http://robot-dreams.net -- rrivera/acm.org (msn) - grafikrobot/aim,yahoo,skype,efnet,gmail
_______________________________________________ Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost
On 01/22/2014 01:46 PM, Belcourt, Kenneth wrote:
On Jan 22, 2014, at 2:09 PM, Rene Rivera wrote:
Does a regular wget https://raw.github.com/boostorg/boost/develop/tools/regression/src/regressio... Does it work with the proxy or without the proxy? Unfortunately no, neither works.
In a browser, does it fail to connect at all or did you get the github 404 Jedi? Your previous message showed a 403 error when you tried to clone which implies you connected to some server and the https certificate was valid (some server being github). I mention the 404 Jedi because the correct url is https://github.com/boostorg/boost/blob/develop/tools/regression/src/regressi... -- Thomas
On Wed, Jan 22, 2014 at 3:52 PM, Thomas Suckow
On 01/22/2014 01:46 PM, Belcourt, Kenneth wrote:
On Jan 22, 2014, at 2:09 PM, Rene Rivera wrote:
Does a regular wget
https://raw.github.com/boostorg/boost/develop/tools/ regression/src/regression.pywork? Does it work with the proxy or without the proxy?
Unfortunately no, neither works.
In a browser, does it fail to connect at all or did you get the github 404 Jedi? Your previous message showed a 403 error when you tried to clone which implies you connected to some server and the https certificate was valid (some server being github).
I mention the 404 Jedi because the correct url is https://github.com/boostorg/boost/blob/develop/tools/ regression/src/regression.py
The valid URL in question is: https://raw.github.com/boostorg/boost/develop/tools/regression/src/regressio... Which gives the "raw" data of the file.. Not the web page showing the fancy source code. -- -- Rene Rivera -- Grafik - Don't Assume Anything -- Robot Dreams - http://robot-dreams.net -- rrivera/acm.org (msn) - grafikrobot/aim,yahoo,skype,efnet,gmail
On Jan 22, 2014, at 2:55 PM, Rene Rivera wrote:
On Wed, Jan 22, 2014 at 3:52 PM, Thomas Suckow
wrote: On 01/22/2014 01:46 PM, Belcourt, Kenneth wrote:
On Jan 22, 2014, at 2:09 PM, Rene Rivera wrote:
Does a regular wget
https://raw.github.com/boostorg/boost/develop/tools/ regression/src/regression.pywork? Does it work with the proxy or without the proxy?
Unfortunately no, neither works.
In a browser, does it fail to connect at all or did you get the github 404 Jedi? Your previous message showed a 403 error when you tried to clone which implies you connected to some server and the https certificate was valid (some server being github).
I mention the 404 Jedi because the correct url is https://github.com/boostorg/boost/blob/develop/tools/ regression/src/regression.py
The valid URL in question is: https://raw.github.com/boostorg/boost/develop/tools/regression/src/regressio...
Which gives the "raw" data of the file.. Not the web page showing the fancy source code.
Here's the wget details in case it helps (with proxy). -bash-4.1$ wget http://raw.github.com/boostorg/boost/develop/tools/regression/src/regression... --2014-01-22 15:09:29-- http://raw.github.com/boostorg/boost/develop/tools/regression/src/regression... Resolving wwwproxy.sandia.gov... 134.253.26.250 Connecting to wwwproxy.sandia.gov|134.253.26.250|:80... connected. Proxy request sent, awaiting response... 301 Moved Permanently Location: https://raw.github.com/boostorg/boost/develop/tools/regression/src/regressio... [following] --2014-01-22 15:09:30-- https://raw.github.com/boostorg/boost/develop/tools/regression/src/regressio... Resolving raw.github.com... 199.27.77.133 Connecting to raw.github.com|199.27.77.133|:443... failed: Connection timed out.
The valid URL in question is: https://raw.github.com/boostorg/boost/develop/tools/regression/src/regressio...
Which gives the "raw" data of the file.. Not the web page showing the fancy source code.
Here's the wget details in case it helps (with proxy).
-bash-4.1$ wget http://raw.github.com/boostorg/boost/develop/tools/regression/src/regression... --2014-01-22 15:09:29-- http://raw.github.com/boostorg/boost/develop/tools/regression/src/regression... Resolving wwwproxy.sandia.gov... 134.253.26.250 Connecting to wwwproxy.sandia.gov|134.253.26.250|:80... connected. Proxy request sent, awaiting response... 301 Moved Permanently Location: https://raw.github.com/boostorg/boost/develop/tools/regression/src/regressio... [following] --2014-01-22 15:09:30-- https://raw.github.com/boostorg/boost/develop/tools/regression/src/regressio... Resolving raw.github.com... 199.27.77.133 Connecting to raw.github.com|199.27.77.133|:443... failed: Connection timed out.
Make sure that https_proxy is set to the same value as http_proxy. I believe that wget is attempting to bypass the proxy for secure connections. -- Thomas
On Jan 22, 2014, at 3:55 PM, Thomas Suckow wrote:
The valid URL in question is: https://raw.github.com/boostorg/boost/develop/tools/regression/src/regressio...
Which gives the "raw" data of the file.. Not the web page showing the fancy source code. Here's the wget details in case it helps (with proxy).
-bash-4.1$ wget http://raw.github.com/boostorg/boost/develop/tools/regression/src/regression... --2014-01-22 15:09:29-- http://raw.github.com/boostorg/boost/develop/tools/regression/src/regression... Resolving wwwproxy.sandia.gov... 134.253.26.250 Connecting to wwwproxy.sandia.gov|134.253.26.250|:80... connected. Proxy request sent, awaiting response... 301 Moved Permanently Location: https://raw.github.com/boostorg/boost/develop/tools/regression/src/regressio... [following] --2014-01-22 15:09:30-- https://raw.github.com/boostorg/boost/develop/tools/regression/src/regressio... Resolving raw.github.com... 199.27.77.133 Connecting to raw.github.com|199.27.77.133|:443... failed: Connection timed out.
Make sure that https_proxy is set to the same value as http_proxy. I believe that wget is attempting to bypass the proxy for secure connections.
Ah indeed, that was my wget problem. So wget can download regression.py: -bash-4.1$ wget https://raw.github.com/boostorg/boost/develop/tools/regression/src/regressio... --2014-01-22 16:31:39-- https://raw.github.com/boostorg/boost/develop/tools/regression/src/regressio... Resolving wwwproxy.sandia.gov... 134.253.26.250 Connecting to wwwproxy.sandia.gov|134.253.26.250|:80... connected. Proxy request sent, awaiting response... 200 OK Length: 41541 (41K) [text/plain] Saving to: “regression.py” 100%[=========================================================================>] 41,541 --.-K/s in 0.09s 2014-01-22 16:31:40 (466 KB/s) - “regression.py” saved [41541/41541] But run.py still hangs here: -bash-4.1$ python run.py --tag=develop --runner="Sandia-Linux" --bjam-options="-j16" --proxy=http://wwwproxy.sandia.gov:80 # Running regressions in /scratch/modular-boost... # Creating regression scripts at /scratch/modular-boost/tools_regression_src... # Downloading regression scripts from https://raw.github.com/boostorg/boost/develop/tools/regression/src... --- http://wwwproxy.sandia.gov:80 Traceback (most recent call last): File "run.py", line 76, in <module> '%s/%s' % (script_remote,src), os.path.join(script_dir,src) ) File "/projects/sierra/linux_rh6/install/Python/2.7.4/lib/python2.7/urllib.py", line 240, in retrieve fp = self.open(url, data) File "/projects/sierra/linux_rh6/install/Python/2.7.4/lib/python2.7/urllib.py", line 208, in open return getattr(self, name)(url) File "/projects/sierra/linux_rh6/install/Python/2.7.4/lib/python2.7/urllib.py", line 437, in open_https h.endheaders(data) File "/projects/sierra/linux_rh6/install/Python/2.7.4/lib/python2.7/httplib.py", line 969, in endheaders self._send_output(message_body) File "/projects/sierra/linux_rh6/install/Python/2.7.4/lib/python2.7/httplib.py", line 829, in _send_output self.send(msg) File "/projects/sierra/linux_rh6/install/Python/2.7.4/lib/python2.7/httplib.py", line 791, in send self.connect() File "/projects/sierra/linux_rh6/install/Python/2.7.4/lib/python2.7/httplib.py", line 1172, in connect self.timeout, self.source_address) File "/projects/sierra/linux_rh6/install/Python/2.7.4/lib/python2.7/socket.py", line 571, in create_connection raise err IOError: [Errno socket error] [Errno 110] Connection timed out
On Jan 22, 2014, at 4:35 PM, Belcourt, Kenneth wrote:
On Jan 22, 2014, at 3:55 PM, Thomas Suckow wrote:
The valid URL in question is: https://raw.github.com/boostorg/boost/develop/tools/regression/src/regressio...
Which gives the "raw" data of the file.. Not the web page showing the fancy source code. Here's the wget details in case it helps (with proxy).
-bash-4.1$ wget http://raw.github.com/boostorg/boost/develop/tools/regression/src/regression... --2014-01-22 15:09:29-- http://raw.github.com/boostorg/boost/develop/tools/regression/src/regression... Resolving wwwproxy.sandia.gov... 134.253.26.250 Connecting to wwwproxy.sandia.gov|134.253.26.250|:80... connected. Proxy request sent, awaiting response... 301 Moved Permanently Location: https://raw.github.com/boostorg/boost/develop/tools/regression/src/regressio... [following] --2014-01-22 15:09:30-- https://raw.github.com/boostorg/boost/develop/tools/regression/src/regressio... Resolving raw.github.com... 199.27.77.133 Connecting to raw.github.com|199.27.77.133|:443... failed: Connection timed out.
Make sure that https_proxy is set to the same value as http_proxy. I believe that wget is attempting to bypass the proxy for secure connections.
Ah indeed, that was my wget problem. So wget can download regression.py:
-bash-4.1$ wget https://raw.github.com/boostorg/boost/develop/tools/regression/src/regressio... --2014-01-22 16:31:39-- https://raw.github.com/boostorg/boost/develop/tools/regression/src/regressio... Resolving wwwproxy.sandia.gov... 134.253.26.250 Connecting to wwwproxy.sandia.gov|134.253.26.250|:80... connected. Proxy request sent, awaiting response... 200 OK Length: 41541 (41K) [text/plain] Saving to: “regression.py”
100%[=========================================================================>] 41,541 --.-K/s in 0.09s
2014-01-22 16:31:40 (466 KB/s) - “regression.py” saved [41541/41541]
But run.py still hangs here:
-bash-4.1$ python run.py --tag=develop --runner="Sandia-Linux" --bjam-options="-j16" --proxy=http://wwwproxy.sandia.gov:80 # Running regressions in /scratch/modular-boost... # Creating regression scripts at /scratch/modular-boost/tools_regression_src... # Downloading regression scripts from https://raw.github.com/boostorg/boost/develop/tools/regression/src... --- http://wwwproxy.sandia.gov:80
Works with this patch to run.py. -bash-4.1$ diff ../run.py run.py 71,72c71,72 < proxy = {'http' : a.split('=')[1] } < print '--- %s' %(proxy['http']) ---
proxy = {'https' : a.split('=')[1] } print '--- %s' %(proxy['https'])
Thanks for the help! -- Noel
On Jan 22, 2014, at 2:52 PM, Thomas Suckow wrote:
On 01/22/2014 01:46 PM, Belcourt, Kenneth wrote:
On Jan 22, 2014, at 2:09 PM, Rene Rivera wrote:
Does a regular wget https://raw.github.com/boostorg/boost/develop/tools/regression/src/regressio... Does it work with the proxy or without the proxy? Unfortunately no, neither works.
In a browser, does it fail to connect at all or did you get the github 404 Jedi?
Got 404 with browser.
Your previous message showed a 403 error when you tried to clone which implies you connected to some server and the https certificate was valid (some server being github).
I mention the 404 Jedi because the correct url is https://github.com/boostorg/boost/blob/develop/tools/regression/src/regressi...
Yes, this url's okay in a browser.
participants (3)
-
Belcourt, Kenneth
-
Rene Rivera
-
Thomas Suckow