perfSONAR User Q&A and Other Discussion

[Alex Moura <>]

On 07/10/2014, at 16:41, Andrew Lake <> wrote:

Hi Alex,

For the host with the web page that won't load, can you send me /var/log/perfsonar/web_admin/web_admin.log?

Sure: 

https://dl.dropboxusercontent.com/u/970090/web_admin.log

Looking at the Apache logs it looks like the cgi script is hanging. web_admin.log will hopefully give a better idea of what it is doing while it hangs. Also, does the host have limited external network connectivity?

It’s open to the wild internet.

My guess is that is contributing (since it checks with a remote lookup service to see if your host is registered) but want to understand better since we thought we protected things like this from happening in such cases.

For the host with the incorrect version in the login message, what happens when you run "/etc/init.d/generate_motd start" and then re-login?

Sorry, I couldn’t try this because when I ssh’d the host, it now shows the correct version somehow (maybe the host need some time to get used to the upgrade):

[07-10-2014 17:20:21 asm@mac ~]$ ssh
's password: 
Last login: Tue Oct  7 17:19:56 2014 from xxxxxx
Welcome to the perfSONAR Toolkit v3.4

You may create accounts to manage this host through the web interface by running the following as root:

  /opt/perfsonar_ps/toolkit/scripts/nptoolkit-configure.py

The web interface should be available at:

https://200.159.254.115/toolkit
----------------------------------------------------------------------

 

Thanks,

Alex

Thanks,

Andy

On Oct 7, 2014, at 2:36 PM, Alex Moura <> wrote:

On 07/10/2014, at 13:45, Andrew Lake <> wrote:

Hi Alex,

On Oct 7, 2014, at 12:30 PM, Alex Moura <> wrote:

Hello, 

After upgrading a pS box from 3.3.2 to pS 3.4, its web interface stopped working:
------------------------------------------------------------------------------

http://200.19.119.101/toolkit/

Gateway Time-out
The gateway did not receive a timely response from the upstream server or application.

Can you SSH into the host?  The gateway timeout could be a number of things. That's what happens when apache doesn't respond. Maybe try restarting apache (/sbin/service/httpd restart) or rebooting the host. 

I did restart the http service and also rebooted the host, but the result is the same. 

Maybe there is another process, - i.e.: cassandra - which the web interface depends on?

Now I noticed there are lots of error messages in the http log like this:

/var/log/httpd/error_log:[Tue Oct 07 15:26:34 2014] [error] cassandra_db [DEBUG] Flush called

The shell is likely the more inaccurate of the two. I believe motd (the file that provides the login message) is regenerated after reboot. Did you reboot this host after upgrade?

Yes, I did.

The filtered log without the "cassandra_db Flush called” errors is available at: https://dl.dropboxusercontent.com/u/970090/error_log.txt

The complete log is available at: https://dl.dropboxusercontent.com/u/970090/error_log_full.txt

Thanks,

Alex

--------
WARNING: At least one of the links in the message below goes to an IP address (e.g. 10.1.1.1), which could be malicious. To learn how to protect yourself, please go here: https://commons.lbl.gov/x/_591B
--------

Attachment: smime.p7s
Description: S/MIME cryptographic signature