Makes sense. I like having pre-configured / pre-filled in templates: that will make it much more user friendly. Having checkboxes and a list (or group) of hosts to apply to would be awesome..drop down box for templates might be cool too. Also, I'm curious - how is device status in the DB set (to 0 or 1)? That seems pretty accurate just for telling me some device is up or down from NMS point of view. Seems like an easy / logical quick win for your alerting framework?
I have about 180 devices and soon around 300 devices in Observium but are still kinda in a 'test' environment. We'd like to replace SolarWinds and their crazy pay for everything nickle dime sales methods with this product. If you'd like I can help test w/ you. If so hit me up off list. I can also volunteer time to help document stuff. Not sure how to integrate to help out there tho. Kenny
Kenny
On Wed, Jan 22, 2014 at 8:48 AM, Tom Laermans tom.laermans@powersource.cxwrote:
Kenny,
Indeed the alerting is still very much in test phase. We don't support alerting on pingable or not yet anyway :[ When most of the functionality is fleshed out the web interface could certainly use some more work. I'm thinking no manual textbox entry of the entities and matching criteria, possibility to prefill input with often-used templates ("Check all BGP sessions", "Check for storage >90%", etc) so ease entry of alerting.
Once it's set up it works for me though, so at least that's a good thing .. ;)
Tom
On 01/22/2014 05:35 PM, Kenny Sallee wrote:
Regarding alerts and status / email - I've had a hard time working with that (please don't judge! It's not that straight forward yet but it's a great product and I'm a happy subscriber).
What I did was created a script that dips into the database and looks at the device status. So I do this basically (again don't judge I'm not a developer yet and still learning I can already see other ways to do this - but it's working so...):
$sql = 'SELECT device_id,hostname,status FROM devices WHERE status = 0'; $count = 'SELECT COUNT(*) from devices WHERE status = 1'; $count2 = 'SELECT COUNT(*) from devices';
$stmt = $dbCon->prepare($sql); $stmt->execute(); $down = $stmt->rowCount();
$stmt2 = $dbCon->prepare($count); $stmt2->execute(); $hosts = $stmt2->fetch();
$stmt3 = $dbCon->prepare($count2); $stmt3->execute(); $t_hosts = $stmt3->fetch();
if ($down < 1) {
........bunch of other logic here...email me if you want it
}
Then some php logic to display via build pretty bootstrap interface to display the count of up and down hosts...if there are any down hosts where status = 1 email goes out. I'm not sure what logic in Observium sets status to 1 (have not looked for it yet)
This has been as accurate as SolarWinds for a customer I work with -- for just knowing if the device is pingable / pollable or not.
Hope this helps, Kenny
On Wed, Jan 22, 2014 at 8:06 AM, Tom Laermans <tom.laermans@powersource.cx
wrote:
On 01/22/2014 04:40 PM, Josh Hopper wrote:
Are there any reporting/analysis functions in the subscription
version? Our CIO is interested in measuring 95% on all of our regional office site links in order to analyze capacity planning.
Adam might have a different answer. There are the metrics that you see. No direct reports.
Observium shows many things if your looking at the right info. It captures traffic statistics for port-channels, ports, VLANs, routing, UDP traffic, TCP/IP traffic, errors, etc.
You can do something with billing, by creating a bill for those ports. That will provide you with historical 95p listings. It's not really reporting, though I'm told billing can do PDF exports somehow. :-)
How mature is the alerting in the subscription version? Is it
possible to configure alerting based on anything other than a percentage of link speed and interface up/down?
Ports measure more than just link state.
Click on one of your interfaces. Notice it tracks traffic, unicast, non-unicast, packet size, utilization, and errors.
We have a number of entities and metrics we alert on. Sometimes it's easy to expand this, in case you need anything specific, sometimes this will take some more effort. You can find a list of metrics we currently can alert on at http://www.observium.org/wiki/Alerting_Metrics_and_Attributes
I have been unable to get any email functionality working in
the community version. Is it supposed to work?
That’s because it’s the legacy system. Last month the alerting system went through like 2-3 updates or something alone.
The community version doesn’t have the nice alerts. Get the subscription.
It may work, or it may not - we don't provide support on that anymore, and have recommended for a few years to keep this disabled due to possible alert mail avalanche... :-)
Regards, Tom
observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
-- Netravine Founder | Kenny Sallee | Kenny@netravine.com | 503.330.2430
observium mailing listobservium@observium.orghttp://postman.memetic.org/cgi-bin/mailman/listinfo/observium
observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium