Comments
-
How are you going to aggregate by table if there are multiple indexes? This might be a bit of a kluge, but you could create a table in the monitored instance to store the data you pull (just insert the selected values along with a date/timestamp of when it was pulled. You can set a threshold to something ridiculously high…
-
You can actually turn off port 8123 and only use the HTTPS port of 8124 if you want. Turn off HTTP traffic in DPA (Port 8123) - SolarWinds Worldwide, LLC. Help and Support will show you how in the <DPAInstallDirectory>\iwc\tomcat\conf\server.xml file. Or, you can try modifying the connector to include scheme="https" like…
-
ChrystalT Love your enthusiasm! Will look forward to "seeing" you on the chat with us! Yes, this should be a pretty awesome discussion - I've worked with Tom for many years and heard Karen multiple time. Together, they bring insight, humor, education, experience, and a natural rapport to audiences.
-
Alex, since this is getting more into the table structure and relations, can you submit a case on this? Perhaps once you construct your query to pull the info, you can post it back to this thread... Note: DPA is NOT to be used for auditing purposes. Our polling occurs once per second, so there will certainly be logins and…
-
If you are using DPA to monitor both instances (the one hosting the primary replica and the one hosting the secondary replica), yes. You can do a report on the read-only instance just on the program in question or databases. Better yet though, you can go into trends for the secondary and drill in through the programs…
-
Great question. The details on the workaround get a little involved and I'd suggest you submit a support ticket so we can help you out with the details. Essentially though, we use a jTDS driver (not MS jdbc driver) for SQL Server connections. There were some fixes put in place to accommodate the multisubnet failover and…
-
I'd go with the second. If you freeze the repo, DPA will keep trying to run transactions and query against it, so it will fail. We do have some retry logic and really nothing you can do will "break" DPA, but it may cause brief gaps in monitoring. With a live dump, that shouldn't impact anything on the repo side as far as…
-
This is looking a bit like a support case for your questions for a couple reasons: * When getting into querying the repo directly, we may have to share data model details * If the data is different between hourly summary and what's seen in the UI, it could be due to averaging the detail data into hourly, but we'll want to…
-
In your alert definition, there should be part of the UI where you specify the instances to run the alert against. Remove the ones you don't want to monitor. Another place to check is the alert group. That can map multiple alerts to multiple instances - it might be configured in there as well.
-
This will likely taking a peak at logs. Recommend opening a support ticket for this one. Submit a Ticket | SolarWinds Customer Portal
-
We run a quickpoll once every second to extract information from SQL Server (see query below). We view this data from the lens of SQL Server, so my question is why would SQL Server be telling us this (s.hostname)? We had another case where a query was being executed from a spreadsheet where the hostname was hardcoded in,…
-
There IS such a page for SolarWinds Core products. In regard to SolarWinds compliance, please see the SolarWinds Readiness statement, https://www.solarwinds.com/general-data-protection-regulation-core-it Additionally, the software products purchased from SolarWinds are products that customers use in an on-premise…
-
Try looking in CON_SQL_NAME table. See if that's what you were looking for.
-
Great questions. First, DPA does not track the time of just one execution unless it is only executed once. This comes down to our methodology and the overhead associated with doing that type of analysis. We track wait times and its associated metadata (user, client machine, plan, etc.) every second. However, we only gather…
-
Do you want them to run more or less frequently?
-
Hey Ed - we currently don't have a way to do this other than individually defining the reports.
-
Check out this post: Change_physical_node_name.docx There is a Word doc in that post. See if that does what you need.
-
Unfortunately, it is not configurable. Regarding the time frame, DPA will look at current day (from midnight to the last full hour of data collection, so 12am to 10am if it's 10:44am). It compares the overall wait time observed for that time frame for today and compares that against the historical average for the same time…
-
Eric, you posted this under the Database Performance Analyzer product. You'd likely want to post this under Server and Application Monitor or another Orion product that can do hardware health... This product doesn't interact with OIDs and MiBs.
-
Since DPA is not an auditing product, we can get you all of the queries that we OBSERVED running against an Oracle instance. Since we poll once a second, there may be quick, infrequent running queries that are missed. DPA being a performance-centric product, we nor you should really care about those. Now, for a complete…
-
Yes. However, here's what you may consider: * Install DPA locally in each customer environment - then monitor whatever they/you want to collect performance data on. * Assuming you will VPN into your customer environments, any latency would be on UI rendering and not on actual data collection (which is more important from a…
-
Can you submit a support ticket on this one? We'll be wanting logs to get the full error stack.
-
What flavor is your repository (Oracle, SQL Server, or MySQL)?
-
This may be possible with how you are drilling in. Try this from the trend view. Go to a single day. Click on the interval drop down and select 1 day. You should see 1 bar representing an aggregate for that day. Now, click on that bar and select the waits tab. Click on CPU, now click on the programs tab. That will show top…
-
We will monitor two resource metrics. One is O/S CPU utilization which is all processors available to the host. The other metric is Instance CPU utilization (processing consumed by SQL Server). You can choose whichever one makes more sense to focus on. Is this what you are referring to or am I missing the mark here?
-
I still see the Download query plan in 9.x version. Are you logged in as a read-only user?
-
If you go to your logs: <install_dir>\iwc\tomcat\logs and open contime.csv in a spreadsheet, filter it by the instance/database ID you are seeing the error in * "select id, name from cond" (against your repo) AND by quickpoll* jobs. We will attempt to execute 600 QPs in a 10 minute period. I consider about 80 - 85% of…
-
What database platform are you monitoring? There can be many reasons stats are not available, but primarily, DPA will only be as good as the system level tables allow. A great example is recompilations of SQL statements - that resets all statistics for that query...
-
First I've seen this kind of request or question. 8 ) First, DPA wasn't designed for this kind of automatic flipping through (more manual driven). However, not sure why you couldn't do what you are suggesting. Would it not be something as simple as linking to:…
-
You are correct in that we do sampling (once a second). We do not trace or audit. It would be ironic if a performance analysis tool caused a performance problem. We choose not to do that. Here is a discussion related to what you are seeing with regards to dynamic SQL using literals. I love the idea of being able to group…
