adam.beedell · SRE/Machine Learning Student · ✭✭✭✭✭

Or SAM or SQL Sentry etc etc

"like increasing disk space, ram, CPU and others." This isnt wrong, but feels very much like a wintel/infrastructurey line rather than a DBA's line. Often you'd need to add an index for example and totally ignore the NPM view of it's RAM as that's misleading

With NTA and NPM i'd say you really dont have the options you need

try: and Status in (1,3,14) or HostStatus in (1,3,14)

You dont need to join, the JOIN will make it slower, you can just do where vm.host.hostname = '${N=SwisEntity;M=Caption}'

Manage nodes -> group by Polling engine -> Select all -> Edit -> check creds box -> change creds

(You can do something like this for the component level if that's what you need, it'd just both be different and represent different stuff)

I am wondering if the issue is because the alert is set to Component instead of Application? ^ This is correct - Doing this at the component level would mean your macro would need to go up to application first

Try dropping the newline

Beaut!

I'm struggling to follow exactly what you're doing, but in general yes you can output multiple numbers or multi-statistic stuff. Either you address them seperately, or concat a line of em together, or FOR XML PATH a table of em together

Havnt seen anyone with the directlink account in use for a while!

- Error confirmed bad, maybe something weird with group 541 - DCOM probably unrelated - You should filter out yor environmental stuff, URL and X-XSRF

is \n the newline? probably want to drop that if so

I havn't used one of these bash(?) ones, but I use the powershell ones a lot and I think it's a simmilar format required. Check any working template for the language you're using You want to do something like Print "Message.X: " and "Statistic.X:" on seperate lines Looks like your script currently is producing a binary…

There's an event on Add, but it'll dissappear after a while There's a "Discovered on" or "-By" field you can use with manage nodes or whatever assuming you add via discovery Oldest ping will bring you back very far depending on your retention settings

Not yet, but this is fairly high on the list of "low priority stuff to figure out"

On a modern dashboard, it's a Table widget, on a classic dashboard it's custom table or custom query

Main Polling Engine - @"vinay.by" has expanded on what I was referring to quite nicely

The alerts get generated from the MPE, should troubleshoot there

If you use the regular alerts widget and filter the page, that does work, but it can be weird based on alerts at wildly different levels (IPSLA and Application for example, no obvious connection) Otherwise there's a bunch of good examples on here for custom alerts queries, you should be able to find them with a quick…

If you do it with IIS it's faster than doing it via config wizard from a downtime perspective

Oh, actually i've been thinking about en masse as Automatic rather than In Bulk. In bulk you could set a property, then Manage Nodes, group by that property, select the value from the left, select all, edit, checkbox for account, change account. That'd do it en masse.

Definately doable with the API, without you're looking at maybe bulk discoveries on auto -> alerts to fill properties. Cant discover based on a property without the API I think

Should be able to get through to customer service guys in like 10 minutes on the phone

You want to call support and talk to the customer service guys (before the tech support guys) License reset takes like <30 mins usually Get logged into the customer portal

@"aycode" you can see this if you look at the alert details page for any alert that's triggered multiple times on the same object. AAT:ID in the url, alert history on the top right showing multiple triggers

I'm not going to look up the detail right this moment, so forgive me if I get a term wrong, but I've made simmilar so feel OK to answer the question anyway You get an Alert(activeIthink?) ID for each config/object pair. So if you want to know if there's a ticket you can reopen, you can reopen the last one with that ID, or…

Bob's correct and yeah this works. You can REPLACE on the message variable to do additional formatting if need be etc

@"wex" Or have em persist a table and update that table with the stored proc, then just read off the bads as before