mnagel

Members
  • Content count

    231
  • Joined

  • Last visited

  • Days Won

    43

Community Reputation

67 Excellent

3 Followers

About mnagel

  • Rank
    Community All Star
  • Birthday July 17
  1. This is a pretty important feature needed and I see no feedback from LM at all. This one seems like it would be pretty simple to add as an option to the report configuration -- please add ASAP. It is hard enough to get folks to pay attention to reports without having them have to go through additional effort to do so. I would much rather see this than whatever effort was applied to rename Devices to Resources.
  2. It would be very helpful to have a button that shows all applicable live alerts for a given rule in the policy editor, especially when deciding whether it is safe to activate a new alert / escalation chain.
  3. More options for alert trigger interval

    And really, what is needed is a general calculation capability on recent data to generate alerts as some items only are important over longer periods of time or when they happen N or M samples. For example, https://www.zabbix.com/documentation/3.2/manual/config/items/itemtypes/calculated
  4. I would like to see an ability to select instances for widgets (and alert rules, etc.) based on other factors than the instance label or glob pattern. It should be possible to select based on an applies-to-like function using instance properties, instance description and instance group (latter I requested separately some time back). A good example for this would be a widget that displays all WAN or DIA links in one place. Selection would be determined either by instance description match (with instructions to the client to ensure those are labeled to cause inclusion) or perhaps by an instance property. I can do some of this with an API-based tool for widget maintenance, but that would not address all places this should be possible.
  5. Cisco Switch ErrDisabled Status on Port

    The one I am using has an ugly side effect -- if you only want to do this and you have LMConfig, you will activate LMConfig since you must define ssh.user and ssh.pass for this, just like what is needed for LMConfig. I recommended to our CSM that LMConfig should require a specific activation property as it is a premium feature that should not be inadvertently activated, but that did not seem to go anywhere. The only option otherwise is to manually edit every CS to adjust the applies-to formula, which then must be updated every time a new version arrives (though I am told better merging is just around the corner...fingers crossed!)
  6. Add Search to Recipient Groups View

    Someone please do this. The UI in general needs to be normalized so all tables can be searched, all objects can be cloned, the same buttons in the same location for the same functions, etc. This one in particular is super annoying once you exceed a page of data (as it would be for any multi-page table).
  7. netflow filter improvements

    I see all the crickets have come to this F/R to hang out. This is a pretty important improvement for using NetFlow for incident research. For example, if you find an IP that is doing a lot of traffic while trying to identify a problem and that IP is harmless, I should be able to filter the harmless IP out of my search as I iterate. There is currently no non-API way to do this. If the filters could be complex with AND/OR/NOT and groups, then it would be much simpler to make use of the data for real world investigations. Similarly, it seems like saved filters are per-user and it would be far more useful if they could be shared across multiple users.
  8. There are times when field values in LM do not match expected field values in integration targets and it would be handy to have an option within the integration to transform data when necessary. Ideally, this would be via the template feature that doesn't yet exist, but it could be done now if the integration data was processed optionally as a Groovy script. A practical example of this is for PagerDuty, where they can accept a severity field. Unfortunately, the values do not match for ##LEVEL## to what PagerDuty expects, so it is a non-starter. If we could execute a Groovy script instead, it would be possible to adjust the values before submitting to the PagerDuty API. I envision this as an option in the integration data field where it would be 'plain text' for the current method and 'Groovy script' for the proposed method.
  9. dashboard sorting

    Please add a display priority field like is available for graphs to allow dashboard order to be controlled other than default lexical sorting.
  10. I am surprised it took me this long to hit this one, but it seems it is still a problem, definitely needs to be fixed. Given the last update here is 2015, seems like it is not being looked at. There are other MSP multitenancy issues I have raised with our CSM repeatedly, perhaps they are all being addressed at once :).
  11. It is currently impossible to create a customized email alert for clients without using a custom email integration, and the custom email integration forcibly disables the ability to support reply commands. I was told (seriously) by support that I should write a script to edit EVERY logicmodule alert message to get the behavior desired. I have opened other requests on getting a real template solution into LM that is sorely needed, but this one is a simple fix that will at least allow adding customized output per client, such as pointers to internal documentation, contact details, etc. Folks do not want to give up the reply command feature to get that and there is no reason they should have to. Please fix this ASAP.
  12. I have frequently run into cases due to many nearly-the-same-but-not-exactly-the-same datasources where alerting should be handled the same and preferably in one rule, but you have to create many rules to achieve that. Multiply that by many different clients and it is very annoying. A good example is reboot alerting and corresponding SNMP failure alerting (via missing data alerts). There are several different datasources that produce that information and I would prefer to have the equivalent of an applies-to function in which the DS/DP could be encapsulated once, then the rule could just say "match for reboot, whatever that means". I would still want to have the device group/device reference separate since those are generally different for each client but those could also be handled by a different applies-to like function, again to be able to manage alert rules much more elegantly and without introducing errors or leaving important cases out by accident.
  13. multiple parallel datapoint thresholds

    Yep as do I. But I generally recoil at changing LM-provided datasources since it makes it much harder to deal with updates. That is a subject of at least a few of my other FR posts and something I am told is actively being worked on :).
  14. Global 'Make Default' for Dashboards

    Agreed -- same issue and this is one of the various areas that don't play well with multiple separate clients (others include monolithic config sections, like default service checkpoints, alert policy and escalation chains). Adding the default dashboard to account/role settings would be very helpful. And probably also solves your problem, @Mosh
  15. multiple parallel datapoint thresholds

    Never saw any feedback on this, and here we are a year later. I just ran into another case where this would be useful (there are many). The current Cisco Stack DS has a State datapoint. That has a builtin threshold for > 4 as warning. Really, this should be "< 4" warning and "> 4" critical, but this is currently impossible. Please make fixing this a priority!