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

The latest update for SWQL studio has coverage for the new port. I'd guess firewalls/network access otherwise

This is the way. I'm also a big fan of adding couple more queries around functionality. 'Does Important_Table_X have an update this hour?" etc

in SWQL you're fairly limited on deletes. Probably works, depends what the object is. I dont think you could or should delete a whole table. Could do whatever in SQL if you had to

Usually super-high frequency stuff ends up being log based in one way or another Either 1) Poll every 10 secs 2) Configure it to send a log, read that. Issue being if it's issue is network based you might not receive it 3) have a script check it every second or whatever. If it fails, output to a log. 3a) have it ask SLW to…

Some options offhand not exhaustive

Like so?

Probably that^ Better to use SWQL if you've got the option to dodge those problems

Try this With logs at any level (or off) and NCM daily backups enabled

https://orangematter.solarwinds.com/2018/06/19/solarwinds-grafana-better-together/ https://github.com/etichy/grafana-swis-datasource https://thwack.solarwinds.com/product-forums/network-performance-monitor-npm/f/forum/63029/grafana-npm-datasource

Is it the certificate thing stopping you from doing SWQL to Grafana?

Date.Time BETWEEN GETDATE() - 7 AND GETDATE() Date.time is looking for a column called "time" in a table called "date" and neither of those exist, you want something like Ex.Observationtimestamp BETWEEN addday(-7, GETDATE()) AND GETDATE() or Ex.Observationtimestamp BETWEEN addday(-7, GETUTCDATE()) AND GETUTCDATE()or…

On 2023.3 here, I cant say I had mine knocked out during the upgrade so not sure what happened to yours there. Hopefully there's a convenient backup!

As a side note, ref with to: "Another limitation on SWQL is that it doesn't support subqueries..." I use SWQL Subqueries every day, totally works

@"matt_rees" Roughly: Custom SWQL on nodes Left Join NCM.Nodes ncm on nodes.nodeid = ncm.corenodeID left join Cirrus.LatestComparisonResults c on ncm.nodeID = c.nodeid WHERE c.diffFlag = TRUE AND c.comparisontype = 4 This probably needs to trigger a comparison so it can read off the changed stuff.

@"matt_rees" Spent a bit of time in the DB earlier, scratch what I said earlier about ncm.configarchive, seems like the fields that look most userful in there arent updated the way you'd think they'd be. I've got an active alert looking at startup vs running which works off the table Cirrus.LatestComparisonResults, I think…

You can, you probably want to add more stuff and not alert on much of what's in appinsight. What you add in depends a lot on what the DB does, best to talk to your DBA if you have one and see what they care about

Had a look through my alerts to try and find what I made before and cant find the thing, found all my change detection stuff turned off, grr. There's a table NCM.ConfigArchive addressable with SWQL, and linking to Orion.Nodes via NCM.Nodes which has enough to build an alert off, albeit not a very easy one to write. I'll…

...Oh dear, well I guess that's the problem Restore that table from backup perhaps?

This probably will string someone then, good troubleshooting

In order of likelihood that you need to do it: Where statements Join statements Join a subquery (in the white box underneath)

In addition to an exit code as Bucweat mentioned, also you need to change the write-host to: Write-host "Statistic: $statistic" Write-host "Message: $message" Where statistic is a int or a string that looks like an int, and message is a string. You can do up to 10 pairs of "Statistic.thing: $thing" too. I've wrote similar…

God I have some much... much... more cursed examples of getting this done. Good stuff. The main need for pivoting I think is the Modern Dashboards Proportional Widget, and as this stuff isnt easiest, it'd be really good to get a version of that widget for ... Portrait data, lets say, or similar.

I've no idea if they're stored anywhere in a file directory, but this suggests its stored direct in the DB, which means that you might be able to fish them out of inactivity or get them back from a partial DB restore

Do you have anything come back from this query? SELECT TOP 1000 AccountID, AssetID, ID, ProjectID, DisplayName, UpdateDateTime, CreateDateTime, AssetType, FileType, FilePath, Persisted, Height, Width, Content, Thumbnail FROM Orion.Maps.Assets

APE issue resolved with excepting windows defender

Attempt 3 website's working normally. One failed install on an APE...

Upgrading at the moment Attempt 1 all installer/config wiz reports are clean. Menu bars blank Attempt 2 all installer/config wiz reports are clean, Menu bars blank, website all messed up. ...

Btw I think the script disclaimer has gotten a bit overzealous, not sure one on every post is ideal "Please note, any custom scripts or other content posted..."

Super weird one, and I dont have the azure bits for any equivalencies. There's some port stuff with 17777/17778 changing toward 17774 that jump to mind perhaps, though not sure how that'd be involved exactly. Very curious how it's solved as you continue

In addition to what Kevin said, which is correct Could also export out (or pull in) with SQL directly, or dashboard from the frontend thing to solarwind's backend without involving a secondary db. or just dash in solarwinds.