SDT management and entry is very time-consuming. As a new user, I just spent a few hours setting SDT for our servers (around 100) and in our previous monitoring system this would have taken about 10 minutes. As a start, create a single screen that shows the SDT for all hosts.
More importantly: the SDT form does not save the correct schedule about 20% of the time and instead saves the item as an \'\'all day\'\' outage. This is bad.
AdvisorSDT Management is being substantially changed in the new UI - unfortunately, its not out in beta yet for hosts. (You can see it on the services tab if you switch to the new UI, however.)rnBut in any event - it should not have taken hours. Thats poor design on our part... Are the SDTs you are setting not aligned with hosts in groups, letting you set SDTs on the group level?
Hi Steve, rnI exaggerated about it taking me a few hours, but it felt like it. I like the cascading SDT feature very much where I can apply SDT by groups and have servers in multiple groups and thus have aggregated SDTs - its really useful. The reason it took me so long is (1) we intended to tune SDT for each server by its network, role and its boot order. The lack of visibility of the aggregate SDT made me give up on that and do a more one-size-fits-all approach and (2) the form often would not save my numbers and instead saved it as an all day outage (using Chrome on Win8.1). RE (1): I may be missing something, but the only way to determine the SDT on a particular server is to right click it, choose SDT and then click the tab to see SDTs affecting this object.
Steve, you bring up a great question with the SDT in alignment within groups. For the OP, I would do this. Configure a custom property for each of your servers that contain some bit of information by that you can then create dynamic groups with. Then set your SDT by group as Steve noted. You might be able to get all of your SDT down to way less than ten minutes!rnThe custom properties is crazy powerful. We even do things like creating a property called contact that has a particular systems owner and contact information in it. Then we include that custom property in all alerts related to that system. We never have to hunt down a system owner this way.rnHope this helps!
