mandevil · Product Manager · ✭✭✭✭✭

This looks to be a built in SPROC in SQL Server. We don't call this from DPA. Can you see other details in the other dimension tabs to determine where it's being run from, who's calling it, etc.? When you are drilled into the SQL statement like you are, toggle through the other tabs across the top to see the DBuser, client…

The answer is absolutely. My recommendation would be to register both instances separately (NOT using the AG listener). Then you will see all activity for both instances including any AG activity on those two replicas.

OK, I believe there are a couple questions in this post. Let's untangle them so they are more bite size. Q. How to exclude things from the charting so that they don't skew the scale so much A. From the trend view, you can click on the SQL hash on the right which will bring up a window. Move the exclude SQL slide to on. A.…

Here is a custom alert you can set up for fragmentation: Custom Alert - Fragmentation From a performance point of view, DPA doesn't focus much on fragmentation because it's not that big of a performance hit in modern data centers. In fact, it could even improve performance in some cases. My intent is not to start a debate,…

We currently only support SMTP. SMS support is a great feature request though. I believe you can send in SMTP format to a device, for example 5555555555@att.net (don't use that though as that's my number). 8 )

Eric, yes, it definitely seems like something is off with the metric as reported. 8 ) Do you mind submitting a support ticket? We can work with you and go after the raw data to see what's throwing it out of whack to far.

This example will be to look at impact over 1 day. Go to trend view, click on the day you want to investigate, from the interval drop down, choose 1 day (that will squish *technical term* the waits into 1 vertical bar), click on that bar. Now go to the machine tab and choose server B (from server A, that will be the…

Unfortunately, not at this time. That sounds like a good feature request though! Check this one out as I think it comes close to what you want (idea expanded upon by GCP).

Is the condition being met over and over again causing a new alert condition to be met? You may look at this one: Custom Alert - Long Running Query (Oracle) Might provide you with the ability to customize it more to your liking regarding the conditions...

A custom metric is limited to a numeric value returned. Having said that, there is already a default metric called Round-trip Time that measure a response from a "SELECT 1" in ms. Are you thinking something along those lines? This would be an indication of network latency though from the DPA server to the monitored…

Here's one you could likely implement and modify: Custom Alert - Blocked Sessions - From Instance It concatenates the select clause. Be aware though that I think our message has a 100 or 200 character limit, so you will likely have to trim it down significantly.

Glen, you should still have that functionality. What do you see when you click into a specific timeframe (let's say drill into 1 hour), click on the SQL statement in question, and look at stats? Execution counts should be part of that data.

Currently this ability does not exist. I like the idea, but we'd have to build additional logic into the product to have say an action based on an alert which can run/email a report. You can submit a feature request here: Database Performance Analyzer Feature Requests‌ (over on the right side is a submit feature request…

Have you checked out task manager when this is occurring? Was the xp_errorlog consuming a lot of CPU? Another process? Was CPU rather highly utilized and this could have been the proverbial straw? To answer your question though, I haven't heard of other cases where CPU utilization was driven that high up by accessing the…

Actually, we only support Sybase ASE...

Check out this document - let me know if that's what you were looking for: Change_physical_node_name.docx

So you will have to first find that SQL within your production instance within DPA. That may require you to set your context more specifically than just overall view of performance (meaning drill into a day on the trend view, set the interval (upper left) to 1 day, and click on that one bar). Then select one of the…

If you check in the iwc.log, are you seeing any SMTP type of errors?

Can you go directly to that query in the trend view and click on "view historical chart"? If you scroll down, you should be able to see a chart with avg execution time over the days being monitored. See what that says. There could have been an anomaly that we're calling out like a blocking situation or ??

Are you wanting to create a custom alert or alert off of the default PLE metric that already comes with DPA OOTB?

kill-9 (love the username!), Are you running this custom alert as in this page? Custom Alert - Long Running Query (Oracle) If you are, you can simply add 'and username <> 'someschema' or other logic into the alert. Pretty simple statement: select 'SID:'||s.sid||', Serial#:'||s.serial#||', Username:'||s.username||',…

Each report is run against an instance. You can run the same report against multiple instances, but you'll have to collate the data manually. One thing you can try is to create a group on the dashboard of instances you want to compare. When you click on the group level, it will show relative information about all of the…

Holger, I think this would be a feature request. I've tried to play around only see that I can sort by a specific column, not re-order the columns.

Alert Levels and Notifications Notification Policy The notification policy can be rather confusing to understand, but is set up this way so that a user can choose for a contact not to be bothered every 10 minutes by alert emails for known issues. The available notification policies are as follows: * Notify when level is…

Not sure who told you that is possible. That would require us to leverage tracing which we don't do (too much overhead on the monitored instance). We will see the effects of the query after the substitution has actually happened (did the execution plan change). We don't peek.

Check this out from the admin guide: SolarWinds Online Help Hope that helps...

You have to be drilled into an hour timeframe. Otherwise, there's a LOT of data to parse through - our attempt to be efficient and as quick as possible.

Is your repo Oracle or SQL Server? I'll see if we have anything close to what you are looking for..

Go ahead and create the custom metric. Then, you can go into any of the instance types it collects for and go into Resources find that custom metric and click on settings. Within settings, go ahead and select custom and uncheck enable data collection and set that as default. This will turn it off by default. Then navigate…

Are you on DPA 12.0? The database is the dpa_repository. The format is db.schema.table. Looks like you are monitoring the instance housing the DPA repo and we hooked one of our own! 8 ) Kind of ironic, eh? The good thing is that we're drinking our own champagne - we've started using DPA to look at ways to make DPA more…