[GeoNetwork-devel] Geonetwork monitor Proposal

Hi,

I have written up a new proposal for adding improved monitoring support to Geonetwork. The proposal is at:

http://trac.osgeo.org/geonetwork/wiki/HealthMonitoring

The basic idea is to be able to set up automated monitoring of the Geonetwork vital statistics so System administrators can have a better idea of the health of a Geonetwork system. This is not the same as the statistics module since the monitoring data is transient and requires an external system to track and store the data. A typical configuration might be a nagios instance pinging the new /monitor/healthcheck url to ensure that the system is functioning correctly and a collectd instance would be calling the monitor/metrics function to check how the system is performing (see if requests are taking longer to handle or if more than normal errors are being logged).

I would love to have some feedback on the proposal.

Jesse

Hi Jesse,

I like your idea. I have often wondered how GN is going. It sometimes
slows down and doesn't allow some sort of functionality which, as an
administrator, I don't know about until someone complains about some
part of the interface not working. (This happened to me last week for
the login service. No-one could login to the interface.)

I also wonder if you could add a "remote monitoring" capability.
Sometimes a GN service is not available because of an Internet
connection. EG. the ISP connection to the Internet may not be working.
However, all local testing seems fine.

If there was an agreement between different GN systems across the
Internet, then one could remotely monitor the other and, if there was a
connection problem, then maybe and email could be automatically sent
from one system to the administrator's email address or the remote
system warning of the situation.

I look forward to seeing your code available in the trunk.

John

On Mon, 2012-03-26 at 09:36 +0200, Jesse Eichar wrote:

Hi,

I have written up a new proposal for adding improved monitoring
support to Geonetwork. The proposal is at:

http://trac.osgeo.org/geonetwork/wiki/HealthMonitoring

The basic idea is to be able to set up automated monitoring of the
Geonetwork vital statistics so System administrators can have a better
idea of the health of a Geonetwork system. This is not the same as
the statistics module since the monitoring data is transient and
requires an external system to track and store the data. A typical
configuration might be a nagios instance pinging the
new /monitor/healthcheck url to ensure that the system is functioning
correctly and a collectd instance would be calling the monitor/metrics
function to check how the system is performing (see if requests are
taking longer to handle or if more than normal errors are being
logged).

I would love to have some feedback on the proposal.

Jesse
------------------------------------------------------------------------------
This SF email is sponsosred by:
Try Windows Azure free for 90 days Click Here
http://p.sf.net/sfu/sfd2d-msazure
_______________________________________________ GeoNetwork-devel mailing list GeoNetwork-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/geonetwork-devel GeoNetwork OpenSource is maintained at http://sourceforge.net/projects/geonetwork

Thank you for the support John. I will keep the list informed of the details as they progress and if you have specific metrics you would like monitored please let me know.

Jesse

On Tue, Mar 27, 2012 at 2:23 AM, john.hockaday <john.hockaday@anonymised.com> wrote:

Hi Jesse,

I like your idea. I have often wondered how GN is going. It sometimes
slows down and doesn’t allow some sort of functionality which, as an
administrator, I don’t know about until someone complains about some
part of the interface not working. (This happened to me last week for
the login service. No-one could login to the interface.)

I also wonder if you could add a “remote monitoring” capability.
Sometimes a GN service is not available because of an Internet
connection. EG. the ISP connection to the Internet may not be working.
However, all local testing seems fine.

If there was an agreement between different GN systems across the
Internet, then one could remotely monitor the other and, if there was a
connection problem, then maybe and email could be automatically sent
from one system to the administrator’s email address or the remote
system warning of the situation.

I look forward to seeing your code available in the trunk.

John

On Mon, 2012-03-26 at 09:36 +0200, Jesse Eichar wrote:

Hi,

I have written up a new proposal for adding improved monitoring
support to Geonetwork. The proposal is at:

http://trac.osgeo.org/geonetwork/wiki/HealthMonitoring

The basic idea is to be able to set up automated monitoring of the
Geonetwork vital statistics so System administrators can have a better
idea of the health of a Geonetwork system. This is not the same as
the statistics module since the monitoring data is transient and
requires an external system to track and store the data. A typical
configuration might be a nagios instance pinging the
new /monitor/healthcheck url to ensure that the system is functioning
correctly and a collectd instance would be calling the monitor/metrics
function to check how the system is performing (see if requests are
taking longer to handle or if more than normal errors are being
logged).

I would love to have some feedback on the proposal.

Jesse


This SF email is sponsosred by:
Try Windows Azure free for 90 days Click Here
http://p.sf.net/sfu/sfd2d-msazure
_______________________________________________ GeoNetwork-devel mailing list GeoNetwork-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/geonetwork-devel GeoNetwork OpenSource is maintained at http://sourceforge.net/projects/geonetwork

Hi Jesse,

I’m very enthusiastic with your proposal. It sounds great !
Please, let us know how it goes
Cheers,

Jean

2012/3/27 Jesse Eichar <jesse.eichar@anonymised.com>

Thank you for the support John. I will keep the list informed of the details as they progress and if you have specific metrics you would like monitored please let me know.

Jesse

On Tue, Mar 27, 2012 at 2:23 AM, john.hockaday <john.hockaday@anonymised.com…> wrote:

Hi Jesse,

I like your idea. I have often wondered how GN is going. It sometimes
slows down and doesn’t allow some sort of functionality which, as an
administrator, I don’t know about until someone complains about some
part of the interface not working. (This happened to me last week for
the login service. No-one could login to the interface.)

I also wonder if you could add a “remote monitoring” capability.
Sometimes a GN service is not available because of an Internet
connection. EG. the ISP connection to the Internet may not be working.
However, all local testing seems fine.

If there was an agreement between different GN systems across the
Internet, then one could remotely monitor the other and, if there was a
connection problem, then maybe and email could be automatically sent
from one system to the administrator’s email address or the remote
system warning of the situation.

I look forward to seeing your code available in the trunk.

John

On Mon, 2012-03-26 at 09:36 +0200, Jesse Eichar wrote:

Hi,

I have written up a new proposal for adding improved monitoring
support to Geonetwork. The proposal is at:

http://trac.osgeo.org/geonetwork/wiki/HealthMonitoring

The basic idea is to be able to set up automated monitoring of the
Geonetwork vital statistics so System administrators can have a better
idea of the health of a Geonetwork system. This is not the same as
the statistics module since the monitoring data is transient and
requires an external system to track and store the data. A typical
configuration might be a nagios instance pinging the
new /monitor/healthcheck url to ensure that the system is functioning
correctly and a collectd instance would be calling the monitor/metrics
function to check how the system is performing (see if requests are
taking longer to handle or if more than normal errors are being
logged).

I would love to have some feedback on the proposal.

Jesse


This SF email is sponsosred by:
Try Windows Azure free for 90 days Click Here
http://p.sf.net/sfu/sfd2d-msazure
_______________________________________________ GeoNetwork-devel mailing list GeoNetwork-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/geonetwork-devel GeoNetwork OpenSource is maintained at http://sourceforge.net/projects/geonetwork


This SF email is sponsosred by:
Try Windows Azure free for 90 days Click Here
http://p.sf.net/sfu/sfd2d-msazure


GeoNetwork-devel mailing list
GeoNetwork-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/geonetwork-devel
GeoNetwork OpenSource is maintained at http://sourceforge.net/projects/geonetwork