cancel
Showing results for 
Search instead for 
Did you mean: 
Create Post
Level 8

is it possible to have switch interfaces monitored to get graphs, but not alert on up/down?

Jump to solution

Hi,

i'm a beginer in Solarwinds and i added a bunch of devices with all theyre interfaces, but now i get tons of emails for each up/down alarm.

is it possible to change the alarming not to act on up/down on some interfaces, but keep them in monitoring for graphing purposes?

Regards,

Ciprian

0 Kudos
1 Solution
Level 12

Just as a point (or half dozen points) of advice for a beginner:

1) defiantly play with all of the built in alerts, but then disable them.

2) If you like the alert, Duplicate it to make it your own, even if you are not going to change it.

3) Prefix all of your alerts with something consistent.  I used our company initials, which is convenient because it  started with an A, which makes is sort to the top of the list.

4) Rename the actions. To make since with the alert type.

     a) So if it's a node down alert named "AAA - Node Down", make the actions like "AAA - Node Down Trigger - Send Email", and "AAA - Node Up Reset - Send Email".

     b) these are just random examples, not specifically even what I use.  Remember the 3 hardest things with developing in IT are "Naming Standards" and "Off by one errors".

     c) You can edit/rename the actions in several ways

          i) Action Manager Tab.  you *will* screw this up as things get complicated, so I suggest avoid using it directly.

          ii) Edit the Alert and you can edit the actions with in it.

          iii) In the Alert Manager tab, you can click on the action or "# actions" link in the "Trigger Action(s)" column, it will take you to the actions manager tab but with a filter for only that alert.

5) I harp on this all the type: Play with custom tags, they are AMAZING!  I think there needs to a an enter section in Thwack just about custom tag usage.

For example, I have a custom tags on interfaces called "Email_On_Critical" and "Email_On_Warning".  You can enter an email address in these custom fields on each and every interface that you want a critical or warning message for when they reach a threshold.  Now every interface can have custom email destination or no email at all, and I only have 2 alerts defined, instead of 100 alerts if I didn't use custom tags.

Your usage of these tags will very a lot, but that is the power of them.  Dive in and give them a try.

6) Defiantly keep asking questions here.  This stuff may look easy, but it's not, it's complicated. This is by far the easiest product for large network monitoring I have used and yes it's still complicated.  I don't know everything, you don't know everything, no one does... except maybe aLTeReGo... he might.

View solution in original post

9 Replies
Level 12

Just as a point (or half dozen points) of advice for a beginner:

1) defiantly play with all of the built in alerts, but then disable them.

2) If you like the alert, Duplicate it to make it your own, even if you are not going to change it.

3) Prefix all of your alerts with something consistent.  I used our company initials, which is convenient because it  started with an A, which makes is sort to the top of the list.

4) Rename the actions. To make since with the alert type.

     a) So if it's a node down alert named "AAA - Node Down", make the actions like "AAA - Node Down Trigger - Send Email", and "AAA - Node Up Reset - Send Email".

     b) these are just random examples, not specifically even what I use.  Remember the 3 hardest things with developing in IT are "Naming Standards" and "Off by one errors".

     c) You can edit/rename the actions in several ways

          i) Action Manager Tab.  you *will* screw this up as things get complicated, so I suggest avoid using it directly.

          ii) Edit the Alert and you can edit the actions with in it.

          iii) In the Alert Manager tab, you can click on the action or "# actions" link in the "Trigger Action(s)" column, it will take you to the actions manager tab but with a filter for only that alert.

5) I harp on this all the type: Play with custom tags, they are AMAZING!  I think there needs to a an enter section in Thwack just about custom tag usage.

For example, I have a custom tags on interfaces called "Email_On_Critical" and "Email_On_Warning".  You can enter an email address in these custom fields on each and every interface that you want a critical or warning message for when they reach a threshold.  Now every interface can have custom email destination or no email at all, and I only have 2 alerts defined, instead of 100 alerts if I didn't use custom tags.

Your usage of these tags will very a lot, but that is the power of them.  Dive in and give them a try.

6) Defiantly keep asking questions here.  This stuff may look easy, but it's not, it's complicated. This is by far the easiest product for large network monitoring I have used and yes it's still complicated.  I don't know everything, you don't know everything, no one does... except maybe aLTeReGo... he might.

View solution in original post

man awesome stuff, just watched the video /classroom with defining and using custom properties, and it's HUGE so simple, so elegant, wow.

now i "just" need to sit down and do the hard part: think and plan

cheers,

this video : VAILABLE:  ON DEMAND  

Expert Series: Streamline your monitoring operation with custom properties 

cnicules​ -

Thanks for linking this.  I'm brand new to Solarwinds as well, so I can use all the education I can get.

0 Kudos

Hi all,

i got more confused with your answers i know it's normal.

My immediate problem with alerts is that now i have "nodes in critical state" beacouse some interfaces are down (it's a office switch and ppl left for home)

So i marked all the interfaces that are down, as "display as unconnected not down" and the "node" is back in green. but i find it confusing and i will need in troubleshooting situations to see if the actual interface is up or down (getting a user call bla bla bla).

from what i see i probably need to change the triggering action of getting the "node" in alarm/warning, and remove somehow the "user ports".

tags is another matter and yes will need to use, as we need to tag ports by at least 20 criteria (criticality, security, customer....)

side question: it's possible to alert on a port going up, not down?

0 Kudos

There is no reason you couldn't create an alert for when an interface comes up. In the alert trigger, just configure it to trigger when interface is up.

If the device is monitored you will get graphs on interface status if interface is selected to be monitored. If you need alerts you can select either for down or up status to send the alert via multiple ways popup alert, SMS, email per example.

Get Outlook for Android

0 Kudos

hi,

to see if i got it wright: i add all the interfaces from the ListResources (up + down), then i twick the alrting to only alert of the interfaces that i want (with custom properties, for example).

Correct?

Thanks,

0 Kudos

You have an alert, probably with a name like "alert me when an interface goes down," that you can disable from the manage alerts screen.  You'd still be collecting all the stats but not trigger an email when the interface goes down.  Later on you can probably modify the alert to make it more specific, like only notify me when an interface goes down on my routers.

- Marc Netterfield, Github
0 Kudos

You can also select the interfaces as "Unpluggable" so that they don't flag as Down when they are unplugged. I do this for edge interfaces that customers still want to monitor for interface traffic and discards etc or customers who don't have UDT.

- David Smith
0 Kudos