Comments
-
Have you checked out this query to generate a report from the repo? Custom Report - Top Plan Advice See if that gets you close to what you wanted.
-
That query is definitely not coming from DPA. I believe gcp has the right answer. If the session is still connected, you can use DPA current monitoring to view both active and idle sessions. You might use it or sp_who2 or a query in SSMS to determine what you want to know.
-
Probably best to open a ticket on this one...
-
Do this, "select id, name from COND" in the repo. Look for tables in the schema for the ones you want to remove that end in the ID pulled from COND. An example would be I get "12, PROD-123\NODE1" as output from the select against COND. I'd go look for any tables with an %_12. If they don't exist, the objects have been…
-
Beyond the day view, I don't believe you can filter the timeslice view like you want to (basically by selecting one of the elements in any of the dimensional tabs across the top). It would be a good feature request though!
-
Not really any documentation on the wait alerts, but it's pretty intuitive. If you still have questions, I'd recommend opening a support case and then we can show it to you in a GTM. Regarding the trending for a specific SQL, the best chart in DPA IMO is the historical chart. On the trend view, select the historical chart…
-
Ok, let me do some digging also - might pull a rabbit out...
-
It will probably take some time now for the system queries to get flushed out of the report...
-
Unfortunately, there's no way I know of to exclude just a user in the pre-defined (canned) reports in DPA. Here's a link for SQL reports which are really just queries against the repo. You could take these and edit accordingly. Database Performance Analyzer Are you looking for a specific report? Or was this just more of a…
-
Could it be that you just don't have any instances trending up?
-
No problem - good luck hunting this down. Are the selects being issues with the nolock clause?
-
These are my personal thoughts as a DBA: The idea of the alert is to notify you that deadlocks have been occurring in excess of the threshold you choose or configure. The analysis of the deadlocks that are occurring could be quite complex involving more than 2 sessions, different impacts, different deadlock priorities,…
-
I really like the idea of doing this, so posted under our custom alert content: Custom Alert - AG Data Loss Alert Feel free to comment or let me know if you fancy it up!
-
OK, I just set this up in a lab env., we sample every 4 minutes for that particular data, so you'd want the alert to mimic that interval. Create a custom alert (single numeric) and query against the repo for whatever instances/AGs you want to alert on. select est_data_loss from con_ag_database Set your thresholds as you…
-
Can't overlay them - you'd just have to spot check two different report outputs. That's what I meant by over/under - literally line up the reports one on top of the other (different browser windows) and eyeball it. I think to get exactly what you want would be a feature request...
-
Do you mind opening up a support ticket? As suspected, this is starting to get messy and we'll need a full set of logs. We'll be looking for the ticket and you can reference this thread. Thanks!
-
Here is a custom report for top SQL by user: Custom Report - Top SQL for Database User Feel free to modify as needed, but it shows the relationship between CONSW, CONST, and CONU (user).
-
Looks like Jamin provided the status on this one - let me know if there's anything else though!
-
Point of clarification - are you talking about the blocking at the session level and the blocked sessions we show as a resource? Scrubbed screenshots may help. Also, if you aren't keen on posting in this forum, a ticket may be in order so you can do a screen share...
-
Ah, you may actually only be attached to one database. What version of DPA are you running and what version of DB2?
-
Going to simple recovery mode will help a lot!
-
Ah, yes, any real-time monitoring is likely not persisted (assuming it's doing more real-time and not keeping historical data).
-
I agree with gcp that a support ticket is warranted at this time. There may be a couple of things happening. With the SQL Express, I've seen where you had to clear out the dynamic ports within the TCP/IP interface config and keep the static one to 1433. Also, like gcp mentioned, there may be a mapping issue. Once you log a…
-
Ah, yes, perfect. Assign one of those licenses to an Azure SQL DB target (or multiple) and it should start collecting. Reply back if you have any issues.
-
"Normal" is likely subjective. 8 ) Let me take a step back. DPA is primarily about the waits. The waits represent your end user or customer pain because something is delayed (waiting). You could even be experiencing high SQL latency, but no additional waits that customers are observing because very little is being…
-
I don't believe this is possible due to our polling methods and how we store the data/statistics. What we do is build wait time data by user, machine, SQL statement, etc. (what we call dimensions). Then we periodically collect statistics for each SQL Statement observed, but we don't divide that up by user or machine. It's…
-
Here is a reference that you might check out. If this were me, I'd limit it specifically to the table in question and do a diff between getdate (current system time) and last_user_update in minutes. If that diff is greater than say between 15 and 30, send me a warning (as a threshold within DPA) and if greater than 30,…
-
What I meant was when we query the execution count from sys.dm_exec_query_stats, we get 4 executions, not 1. I'd recommend going after the raw data to validate this (take DPA out of the picture). For more assistance with that, can you submit a support ticket from your portal?
-
Thank you for the post! That's awesome! Here is the beta site for DPA: Database Performance Analyzer Beta I'll also let the PM know of your interest - remember, beta should be considered throw away once you are done with it, so you might want to stand up another DPA environment and double monitor dev for a bit (with such…