Closed ghost closed 2 years ago
It sounds like there is still a route in place for reporting.
You'd mentioned you tried opscode-manage uninstall
- was this the command used? I think opscode-reporting-ctl uninstall
is what would have been needed, optionally followed by 'opscode-reporting-ctl cleanse' to purge all reporting data
I think what's happening is that the nginx upstreams are still defined even though the service is not removed. If you take a look in /var/opt/opscode/nginx/etc/nginx.d
you should see some files matching '20-reporting_*.conf'. Those are providing the routes and upstreams for the reporting service.
If you delete them and chef-server-ctl restart nginx
that should cause reporting requests to 404, which chef-client expects if reporting is not installed.
Hi,
of course I used opscode-reporting-ctl uninstall
The directory you told me was empty - but I found another one:
/var/opt/opscode/nginx/etc/addon.d
[root@inf-c23fef1b addon.d]# ls -al
total 44
drwxr-x---. 2 opscode opscode 4096 May 3 12:58 .
drwxr-x---. 5 opscode opscode 4096 May 4 20:44 ..
-rw-r--r--. 1 root root 300 May 3 12:58 20-reporting_external.conf
-rw-r--r--. 1 root root 539 May 3 12:58 20-reporting_internal.conf
-rw-r--r--. 1 root root 96 May 3 12:58 20-reporting_upstreams.conf
-rw-r--r--. 1 root root 466 Dec 20 14:37 30-opscode-manage_external.conf
-rw-r--r--. 1 root root 239 Dec 20 14:37 30-opscode-manage_internal.conf
-rw-r--r--. 1 root root 188 Dec 20 14:37 30-opscode-manage_upstreams.conf
-rw-r--r--. 1 root root 122 Nov 30 21:09 40-oc_id_external.conf
-rw-r--r--. 1 root root 90 Nov 30 21:09 40-oc_id_upstreams.conf
-rw-r--r--. 1 root root 754 Nov 30 21:09 README.md
Killing those files there solved my issue.
+1 reporting module just broke yesterday - just switched the old SSL cert for a new one. uninstalling opscode-reporting-ctl & cleansing - reconfigure - restart didnt helped. removing 20-reporting*.conf file from /var/opt/opscode/nginx/etc/addon.d - did the trick.
Hi Guys..
I had the same problem and I fixed in this way.
Reason: Authentication problem with RabbitMQ. I saw the error in /var/opt/opscode/rabbitmq/log/rabbit\@localhost.log
Steps:
ln -s /opt/opscode/embedded/bin/erl /usr/bin/erl
/opt/opscode/embedded/bin/chpst -u opscode -U opscode /opt/opscode/embedded/bin/rabbitmqctl add_user chef2 mypassword
/opt/opscode/embedded/bin/chpst -u opscode -U opscode /opt/opscode/embedded/bin/rabbitmqctl setpermissions -p /analytics chef2 "." "._" ".*"
3.Changed username and password for RabbitMQ into the following file. /var/opt/opscode-reporting/opscode-reporting/etc/sys.config
For me this worked but I will lose my changed after a chef-server-ctl/opscode-reporting-ctl reconfigure. I suggest to override Chef attribute in chef-server.rb
I just had the exact same thing occur. I was getting 502 errors on my clients and the server (whose logs are absolutely horrible to tail with chef-server-ctl tail
) and Connection refused while connecting to upstream
. I was also getting a few random solr errors I think.
Simply running rm -f /var/opt/opscode/nginx/etc/addon.d/20-reporting*; chef-server-ctl restart;
worked for me.
I had to delete those *.conf
files and then run chef-manage-ctl reconfigure
before the web server would work again.
Create a cleanup reporting if unused task as a fallout action item from this. Reporting is currently EOL.
oc-reporting is EOL
Hi,
I've installed opscode-reporting a while ago but it never worked (for me). Today I decided to uninstall this add-on with
opscode-reporting-ctl uninstall
,chef-server-ctl reconfigure
chef-server-ctl restart
and finallyyum erase opscode-reporting
.I saw that some files where remaining to I deleted them:
rm -Rf /var/log/opscode-reporting
,rm -Rf /opt/opscode-reporting
,rm -Rf /opt/opscode-reporting
,`/usr/local/sbin/refresh_reporting_matviews
Since then starting any chef-client is delayed and logs on client side:
On server side I'm seeing this:
Indeed there's no port 10010 listening on the chef server. Here're my service list, configuration and port bindings:
My chef server is running on:
The clients are mixed from CentOS 5 to CentOS 7 with different versions of Chef 12 - but they're all experiencing the same issue.
I also tried reinstalling opscode-reporting - but this did not solve the issue.