Community

Hi everyone,

Currently i am trying this cloud software, but got issue with the monitoring page.

The monitoring web address is http://monitor.<sitename>.nimbula:82

My problem is that the graph monitoring does not appear (graph.php). I've tried IE, chrome, and mozilla, no difference. The option load image is enabled already.

If anyone has input on this issue, kindly advice.

Thanks!

suiti.chen@yahoo.com
Hi everyone,

Currently i am trying this cloud software, but got issue with the monitoring page.

The monitoring web address is http://monitor.<sitename>.nimbula:82

My problem is that the graph monitoring does not appear (graph.php). I've tried IE, chrome, and mozilla, no difference. The option load image is enabled already.

If anyone has input on this issue, kindly advice.

Thanks!

Mr Chen,
When you tried to access the monitoring page, what kind of error did you get? Did you replace <sitename> with the actual name you used in your site installation? Also, please make sure that your DNS server is updated to be able to resolve <sitename>.nimbula to a valid IP address.
Regards, Bo.

Hi Bo,
Yes, i have replaced the sitename with the name i specified on site.conf. The DNS resolve works fine because the other informations (texts) are loaded successfully, only the image (graph) cannot be loaded.

The weird thing is, the CPU percentage and other value is still 0%, though i already run 1 or more instances..

Thank you.

Edited suiti.chen@yahoo.com (Jan. 25, 2012 17:20:49)

It would help greatly if you could take a screen capture and attach it here. How were you able to figure out the CPU load when the graphs could not show up in your browser?

Regards, Bo.

bli@nimbula.com
It would help greatly if you could take a screen capture and attach it here. How were you able to figure out the CPU load when the graphs could not show up in your browser?

Regards, Bo.

Hi Bo,
It's written on the description on the left pane of webpage.., please find the sample attached..Thanks.

Attachments:

Unfortunately the attachment you put here doesn't give me much idea on what went wrong. If you can instead send me a screen capture in image format, it probably would help me better. In addition, the following information would be very helpful too:
- number of clusters in your site and number of nodes in each cluster
- log files (/var/log/messages) collected from each node. You can do this by ssh to the nodes and copy them to a central place.
From the page you sent, the site is installed on a private network, so I guess I wouldn't be able to access it to help debug the issue. If you can collect the data I mentioned above, you can send them directly to my email (bli@nimbula.com).

Thanks, Bo.

Hi Bo Li,
Currently i only have 1 cluster with 3 node server (1 seed node total 4)
The /var/log/messages size is to big to be attached via email..I've sent to your email, the monitor page in image format, and also the site log (nimbula-get-logs --site-logs) hopefully it can help to assess the issue.

Thx a lot

From the logs you forwarded to me, it looks like the DNS name queries for the monitor daemon hosts were failing. If the monitor daemons were not running, the monitoring data would not be collected, which could explain why the graphs didn't show in your browser.
The logs you collected were not very complete, so we need more information to diagnose the problem. Please try the following steps:
1. ssh to one of the cluster nodes
2. ping monitor1.192.168.131.0 (most likely this would fail)
3. ssh 192.168.131.4
4. sudo /etc/nimbula/bic/bdns_1 status
5. If the status is "running", then 'sudo /etc/nimbula/bic/bdns_1 stop' to restart the DNS server 1.
6. Check the DNS status again as in step 4.
7. If the DNS server restart is successful, repeat steps 3 to 7 with 192.168.131.3 and bdns_0 to restart DNS server 0.
8. Repeat step 2 to see if you can access the monitor daemon.

If you encounter any failures in the process, please collect the following information by issuing the commands:
$ sudo nimbula-test-services | tee service.log
$ sudo nimbula-get-logs (to collect logs)
Forward the logs you collected to my personal email.

Thanks, Bo.

bli@nimbula.com

From the logs you forwarded to me, it looks like the DNS name queries for the monitor daemon hosts were failing. If the monitor daemons were not running, the monitoring data would not be collected, which could explain why the graphs didn't show in your browser.
The logs you collected were not very complete, so we need more information to diagnose the problem. Please try the following steps:
1. ssh to one of the cluster nodes
2. ping monitor1.192.168.131.0 (most likely this would fail)
3. ssh 192.168.131.4
4. sudo /etc/nimbula/bic/bdns_1 status
5. If the status is "running", then 'sudo /etc/nimbula/bic/bdns_1 stop' to restart the DNS server 1.
6. Check the DNS status again as in step 4.
7. If the DNS server restart is successful, repeat steps 3 to 7 with 192.168.131.3 and bdns_0 to restart DNS server 0.
8. Repeat step 2 to see if you can access the monitor daemon.

If you encounter any failures in the process, please collect the following information by issuing the commands:
$ sudo nimbula-test-services | tee service.log
$ sudo nimbula-get-logs (to collect logs)
Forward the logs you collected to my personal email.

Thanks, Bo.

Hi Bo,
Both dns service status is running...i have restart the service as per steps above just to make sure.
The ping to monitor1.192.168.131.0 reply successfully as well..although there is a gap for the first reply...(attached in this post)

I will still send you the other logs from nimbula-test-services and nimbula-get-logs just in case...

Thanks for your guidance so far.

Edited suiti.chen@yahoo.com (Jan. 30, 2012 00:49:31)

Attachments:

The steps you performed and services log you forwarded to me all look good. Do the graphs still not show up for you? Please do forward the syslogs you collected to my email, so we can dig into more.

Thanks, Bo.