mandevil · Product Manager · ✭✭✭✭✭

Unfortunately, it's not us seeing the statement as different, it's the database engine that's assigning a different hash/handle that we rely on. If they are using dynamic SQL generation, what they may want to do is look into using bind variables (that's more from a generic development perspective). I'm not familiar with…

Can you open a support ticket from your portal? I'm thinking this will require logs...

As long as the browser you are using has access to the host DPA is running on, you should be fine just hitting the url. Something like http:\\<hostname or IP>:8123 should get you to the UI for DPA. If you are running DPA within your company's network, you will likely have to be on the same network (so VPN if remote). To…

That was one of the most commonly requested feature additions by our customers during a recent CAB (customer advisory board) meeting. So, great minds think alike. 8 ) If you check out What we are working on now you will see that the file I/O metrics is the top item on our road map. Coming soon to a theater near you. 

Here is a post that may help with some reasons you get a negative hash and no SQL text. Negative Hash values in DPA - SolarWinds Worldwide, LLC. Help and Support DPA relies on what SQL Server provides it...

This one is likely better as a support case instead of a Thwack community post. Thinking they'll need logs on this one.

I'd recommend opening a support case for this one. Odd that it happened right after the licensing issue. Do you have the VM license checked in license manager for this Sybase instance?

Alex, since DPA is a standalone product at this time, best practice would be a separate box for the installation of the DPA server and its repository database. This is especially true if there are plans or potential for expansion. No pitfalls that I know of. 8 ) Are you familiar with wait-based analytics? If not, check…

Have you checked this script out? You might need to tweak it a bit for your exact purposes: ss hourly avg metric SQL Also, we summarize resource metrics, but if you want them within the past 30 days, you should be fine. You didn't specify your repo version, so I just included the one for SQL Server...

DPA has a metric called Round Trip Time that will ping each of your registered instances per the set interval of the metric. This basically does a select from the database which should spend next to no time in the database itself (so a proxy network latency metric). Note though that this will only "watch" your network…

What version of DPA are you on? If more recent version (9.1+), you can view historical data collected while licensed, no issues. You can validate/test this out by unlicensing a non-critical instance, then go back to the main page and still click into that instance to see data that was collected while licensed.

Just created this out on Thwack... Standby/DataGuard Alert

Here is one I've setup previously. You can use the SQL to create either a custom alert or metric. Nothing canned in the product currently. select ( SELECT MAX(SEQUENCE#) LOG_ARCHIVED FROM V$ARCHIVED_LOG WHERE DEST_ID=1 AND ARCHIVED='YES' ) - ( SELECT MAX(SEQUENCE#) LOG_APPLIED FROM V$ARCHIVED_LOG WHERE DEST_ID=2 AND…

First, the scale is in ms (milliseconds) and not %. That is an indication of latency before a commit is confirmed. Second, if you go back to the trend view and click on the Waits tab, you will likely see Log File Sync as one of the waits you are experiencing. Click on the wait link on the right and it should display a help…

When you say email configurations, do you mean for contacts for alerts and notification or do you mean the SMTP server you have configured for emails to go through?

Here is an article discussing this very thing. I'd advise caution making edits directly to the database as there may be unintended consequences. Two "Node Details" system view If you don't want the DPA integration any longer, you can delete the integration from settings...

This was a known bug in 9.0.X of DPA. There was a change in the parameter type for that sproc that wasn't caught. Any possibility to upgrade to 9.1 where this alert was fixed? Sorry for any inconvenience!

Hany, DPA is definitely more of a performance analysis tool rather than an availability or uptime reporting tool. Having said that, you could use some of the metrics like round trip time to provide an idea of availability. As long as the instance being monitored is up and our monitor is up and running, then you'll see a…

Would something like this work? Custom Report - Blocking Report

Can you try from another browser? Wondering if it's just rendering odd in the one you are currently using...

There may be a couple of limitations with this request: * We are connected to the Oracle instance being monitored, not the ASM instance * We can currently plot a custom metric for one value at a time Given these limitations, can you query the v$asm_diskgroup view from the monitored instance? If yes, we should be able to…

So like many database questions, this will be answered with an "It depends". You can host the new repo on an instance with an existing repo. However, watch for resource contention using this multi-tenant approach. You may even want to monitor the shared instance with DPA to see what kind of impact each has on the other...

Can you login to the instance experiencing the issue as the same user that DPA users for monitoring and run an explain plan for the query in question? Any errors reported doing that?

Here is one way: Step 1: Set up an alert for tempdb filling up. Grab the script you feel most comfortable with to check tempdb used space (or free space if you like to look at it from that perspective). There are a lot out there and some are % based, some are byte based, so don't want to assume anything here. If you get an…

The issue we run into is that the database engine itself will assign a different sql_handle or sql_id if anything in the query changes. Really the only way to bridge this is to name the hashes as discovered with similar naming conventions so you can easily recognize when queries are related. How to recognize and do this…

I don't see a way to exclude that parameter in the canned alert. I'd recommend a custom query, but since we're doing difference detection, that's in the source code (java). Thinking you will have to live with this notification (maybe setup a rule to have it put in another folder to remove the noise?).

Jason, have you tried renaming the SQL hash from the UI? From the trend page or timeslice tab, you can click on the name represented on the right hand side and either remove the name altogether (that will display the hash again) or use a different name. If you are thinking of editing the repository directly to do it en…

Lots of questions, but I'll start out by stating I suspect something environmental on your end. * Is your DPA service or process actually running? Can you telnet to the servername/IP:port for DPA? * Is your DPA configured to only use HTTPS connections via the browser? * Can you hit the DPA UI from the server it's actually…

mbird​, the feature I think you are looking for is CLEAN_DAYS_OF_DETAIL which will allow a longer history to be displayed in trends. The default is 30 days. The one caveat is that this will require additional storage for this instance as about 90% of our consumption is in the detail.

Can you confirm that instance is licensed?