Have questions about how to best use LogicMonitor? LogicMonitor experts are live to explain best practices and answer your questions. There are no planned topics as this webinar is a true office hours. Bring your coffee and questions!
Please consider giving us feedback.
Q&A Transcript:
[2:00] Why do some of the new F5 DataSources target the "F5Active" category? There is no PropertySource or SysOID map for that category. At least none that I can find. Something similar with the brand new DS suite for "Microsoft_Windows_Cluster_" which all use Categories that don't have a PS yet. F5_BigIP_VIP_Cert Sorry, wrong order F5_BigIp_SSL...Thanks.
lmLocator: CFYHLR ?
[6:04] I would like a quick overview of monitoring several related windows services and dependancies that represent our customer facing services. For example, server1 - SQL service, server2 - Web service, server3 - URL, server4 -etc...... Would you use a LM "Service" to reflect these components
[10:53] Nice! Thank you. - I'm told that websites cannot be included in these "Services" True?
You can also create a script DS that uses selenium to do synthetic transactions in your service.
GRRRR
That'l work!
[15:15] relating to the LM services question, we have a scenario where we have a custom datasource that has many instances. these instances are duplicated on multiple servers and we only care if the instance is alerting on all servers. the instance names are the same across each node. is there a trick within services so that we can add 1 service and have each uniquely named instance act as its own service, or do we need to speficially create 1 service for each individual instance name? yeah thats what i found, thanks
TY!
Very good.
[16:46] I am really interested in the Mapping functionality, however we are almost 100% AWS cloud deployed. It does not appear that the Mapping services is designed for cloud resources. Is there anything on the roadmap to allow Mapping to see the relationship between a collection of AWS cloud resources? OK, thank you
[18:18] If monitoring using WEBSITES" I am told my license is limited to 100 sites. If we add HTTP_Page to a device is there a limit to sites.pages?
[20:46] Can you demonstrate how you would decide to enable Dynamic Thresholds to suppress alerts that might occur at predictable times, and verifying after the breaking-in period? I've been having difficulty knowing if they are successfully implemented. My example is for memory usage metrics. Very helpful, this conversation clarifies point from the documentation
[29:07] More about Services... Could you talk / show a bit about what permissions are needed to be able to create and/or manage Services? We have a user (who is also on this session but I'm asking this for them) who has full Manage --> All for Resources on one of their Roles, but they can't see any Metrics listed on a Service which has completely valid Members. Even more odd is how someone with a full Administrator Role can see all the Metrics in that same Service where they can't see a single one. Is there something more than Resources Manage --> All needed? (This might be something we need to handle with LM Support. It just came up recently and we haven't had a chance to do so yet.) Yep - View modules is on already. No sweat - thanks for pondering it!
[33:46] OK... Here is a challange for you: How to monitor a website that first does a redirect to an SSO login before coming back to the page.
Selenium!
I followed Groovy hints here and got great results for our SSO apps! https://www.logicmonitor.com/blog/is-this-application-down-or-is-it-just-me
Also good stuff here: https://www.logicmonitor.com/support/services/adding-managing-services/executing-internal-web-checks-via-groovy-scripts
[36:20] I have several switches that serve important links (trunks or servers) that I want to know when they go down. However, the same switches also host workstations and don’t need to be alerted if they port is down. There are more workstations on the switch than important links. What is the best practice to not get alerts from the workstation interfaces?
use of the property "interface.description.alert_enable" lets you set specific interface descriptions that you only want to alert on for status.
https://www.logicmonitor.com/support/monitoring/networking-firewalls/interface-status-alerting
[39:55] When using websites, where there may be several steps added...
if any one of the steps fail, we get alerted, but is there a way to show the specific step that failed?
example: I want to test for steps:
https://site.company.com/path1
https://site.company.com/path2
https://site.company.com/path3
if any path fails I get alerted, but we need to go look to which path failed.
I’ve wanted this too - I’ve set up multiple web checks for the same site 
Wait - ##DETAIL## says "will include the URL for the step that failed"
##DETAIL## is in my alert template, but it does not seem to work. 
can we have alert use a variable/token to provide the actual failed path..?
[43:50] I know it isn't a question but I want to lament that text cannot be displayed outside of "Poll Now". I want to return a number but also some related text. For example, how many people accessed a file and alert if the number is greater than 0, but the engineer responding to the alert would also like to have info such as what action was taken.
If it's a custom script, you could fire the text into an Event Source right?
EventSources are less routable from a AlertRule perspective.
thanks again
[48:06] For Cisco AnyConnect, can LogicMonitor poll the connected users during a polling period. It can pull the number of AnyConnect sessions, but just the number of users, not who at this time.
[50:27] how do we add that token to the alert template?
[52:16] We had a user ask about getting Website Response text to be available in an Alert text that's coming from an Overall Alert. (They don't get nor want to get Alerts from individual Checkpoints.) Is there any trick that would allow the Response (for example ##WEBSITERESPONSE## token mentioned on https://www.logicmonitor.com/support/logicmodules/about-logicmodules/tokens-available-in-datasource-alert-messages#websites) OR can you think of anything that would send only a single Checkpoint Alert message even if multiple Checkpoints are configured for that website? (I had written this question before we started talking about the previous similar question) It's similar but not quite the same I think. I'm thinking of making a Feature Request that would make (for example) the most recent Response text available. Hmmm - I did poke at that, but maybe I didn't come up with the right combo. Thanks!!!!
