perfSONAR User Q&A and Other Discussion

[Shawn McKee <>]

Hi Terrence,

I think it would be easiest to just follow Andy's instructions to get things functioning again.

Also, 3.4.2 is now out.   After you run Andy's commands you should also do:

yum clean all

yum -y update

reboot

I think this should get you setup and operating correctly.   Do check your disk space usage and if needed, clean up /var/log/perfsonar

Let me know if you have questions or hit any problems.

Thanks,

Shawn

On Tue, Mar 24, 2015 at 7:55 AM, Andrew Lake <> wrote:

Hi Terrence,

Looks like somehow your postgres user got removed. Are these new installs or installs that have previously worked?

The following should clear-up your problem but it will delete any data you had previously collected:

rm -rf /var/lib/pgsql/data/*
/opt/perfsonar_ps/toolkit/scripts/system_environment/configure_esmond
/opt/perfsonar_ps/toolkit/scripts/system_environment/configure_regular_testing
/etc/init.d/httpd restart
/etc/init.d/regular_testing start

If you don't care about losing the measurement data the above should do the trick. If you don't want to lose the data, I'll need a couple files so we can pinpoint exactly what is out of sync. Specifically I'll need /opt/esmond/esmond.conf and /var/lib/pgsql/data/pg_hba.conf (and if the latter does not exist that's significant information as well).

Thanks,
Andy

On Mar 23, 2015, at 6:05 PM, Terrence Martin <> wrote:

> I think the issue I have is different, but it seems that there is certainly an issue there.
>
>
>  File "/opt/esmond/lib/python2.7/site-packages/django/db/models/sql/compiler.py", line 839, in execute_sql
>    cursor = self.connection.cursor()
>  File "/opt/esmond/lib/python2.7/site-packages/django/db/backends/__init__.py", line 326, in cursor
>    cursor = util.CursorWrapper(self._cursor(), self)
>  File "/opt/esmond/lib/python2.7/site-packages/django/db/backends/postgresql_psycopg2/base.py", line 182, in _cursor
>    self.connection = Database.connect(**conn_params)
>  File "/opt/esmond/lib/python2.7/site-packages/psycopg2/__init__.py", line 164, in connect
>    conn = _connect(dsn, connection_factory=connection_factory, async=async)
> OperationalError: FATAL:  password authentication failed for user "esmond"
>
> I suspect that during an update a user was added to the services but some part of the system was not updated to be in sync. eg. when they changed the login user from root to admin.
>
> Terrence
>
>
> On 3/23/2015 2:05 PM, Remsik,Robert wrote:
>> I actually just solved this issue not but a couple hours ago!  The below
>> clause helped me out.  I just had to edit the cassandra-env.sh using the
>> suggested values and everything seemed to appear.
>>
>> Give it a try and see if that helps you.
>>
>> http://www.perfsonar.net/about/faq/#Q68
>>
>> Cheers,
>>
>>
>> On Mon, 2015-03-23 at 13:53 -0700, Terrence Martin wrote:
>>> Hi,
>>>
>>> Neither of my Perfsonar systems, the bandwidth, or the latency are
>>> currently able to properly report the data. Some stuff is running.
>>> However in both cases when I attempt to get results I get the
>>> following in the case of the bandwidth host.
>>>
>>>
>>>
>>>
>>> Click a hostname to see a graph of the results
>>>
>>> Grey text indicates inactive test (tests that have not run in the last
>>> 7 days)
>>>
>>> Error retrieving test data: undefined
>>>
>>> Or in the case of the latency
>>>
>>> Error: No Measurement Archives available.
>>> 500 INTERNAL SERVER ERROR
>>>
>>> It seems clear to me the issue is some disconnect between the data
>>> collection and storage. However it is not clear to me where this
>>> broke.
>>>
>>> Maybe it is because something broke during the auto-upgrade, but I am
>>> not certain how to get the system to re-initialize as it were.
>>>
>>> Any suggestions on how to proceed would be greatly appreciated aside
>>> from a complete re-install.
>>>
>>> Terrence
>>>
>