darichar ✭✭✭✭✭

I am checking with our dev team on this, but I believe this is already built in as an option. I will get back to you once I understand it better and the exact setting.

You will probably want to open a support case so we can get your log files and take a look. Alternatively, send your log files to me @ dean.richards@solarwinds.com.

Implemented in DPA 12.1. See DPA 12.1 Release Notes - SolarWinds Worldwide, LLC. Help and Support for more info.

At the current time DPA can exclude queries that execute against objects owned by SYS, but not by schema. I like the request and hope others vote it up as well. As an alternative you can exclude individual queries from Table Tuning. Just click on that query from anywhere and it will take you to Query Performance Analyzer…

Unfortunately, the data loss metric for the replica is queried as the page loads/refreshes and is not persisted in the database currently. You could setup a custom metric that queries the data from the sys.dm_hadr_database_replica_states using the last_sent_time (last time data was sent from primary) - last_received_time…

Is it the same thing, i.e. this metric is the possible data loss from the primary to secondary replicas? If you know of another number that's available, where does it come from?

There are several tables that store AG information, and they all start with "CON_AG". The data loss metric you mention in stored in the CON_AG_DATABASE table in the EST_DATA_LOSS column. There are also columns like SYNC_STATUS, HEALTH, and IS_FAILOVER_READY that can give some of the other information as well. This table…

Yes, that table is cumulative. You would need a job that periodically saves the samples and writes the required data to a table. Then, you can query that persisted table to calculate deltas. SQL Sentry already handles this for you, so you could also retrieve the data from the SQL Sentry repository. Here’s a link to Part 5,…

You would need to change the thresholds listed to 5 min or 300 seconds, but otherwise the custom alert should work as is.

By default, DPA collects the amount of total time for blocking, but you will not find blockers that lasted more than 5 min. I would suggest setting up this custom alert (+) Sessions Blocking for More Than X Minutes - Custom Alerts - Database Performance Analyzer - THWACK. It will notify you when a session has blocked for…

If you only need to monitor the primary node in an AG, add it using the AG listener and DPA will follow it around as failovers occur. If you also want to monitor a read only node, add it using the listener as well but then in the JDBC Properties field under Advanced Properties use ApplicationIntent=ReadOnly as discussed…

Give this a try as well and let me know your thoughts: thwack.solarwinds.com/.../3441

I'm assuming you want to change the Performance Counter so it excludes those waits. You could replace the performance counter with a SQL Server query something similar to "select sum(wait_time_ms) from sys.dm_os_wait_stats where wait_type not in ('CXPACKET', 'CXCONSUMER')". If you are after something else, let me know.

It looks like that alert could tell you when the size of an object exceeds a threhold. Here is a series of 2 alerts, 1 for collecting historical data, and 1 for calculating the growth - Collector Alert that never fires, it just collects data: (+) SQL Server Table Size Collector - Custom Alerts - Database Performance…

Hi @"shantanu167" , here is a custom query that should do something close to what you're looking for: (+) Resource Metric Report (Min, Max, Avg) - Custom Queries - Database Performance Analyzer - THWACK. As mentioned in the script, the CON_METRICS_TEN_MINUTE_X, CON_METRICS_HOUR_X, and CON_METRICS_DAY_X tables offer various…

I might suggest at least 2 monitoring services in each domain, where each pair can provide load balancing and HA for each other.

The Custom Queries section for DPA in Thwack may not provide the exact answers, but the articles may provide you with a starting point for where the data is located: thwack.solarwinds.com/.../custom-queries

Not directly, but I have a customer do the following. You can configure custom properties within DPA and then assigned them to monitored instances. That custom property could then be included in the subject or body of the email and people setup outlook rules to handle the emails accordingly. For example, if the subject…

If an instance is stopped, alerts will not execute against that instance, so this alert is probably not going to tell you want. What is your end goal here, to know when a monitor stops? If so, you could plug your query into a user defined alert within DPAIM. It's hard for a monitoring product to monitor itself. Also, if…

DPA monitors whatever you point it to. For AGs this is typically the AG listener and then DPA will monitoring the primary and follow it around as it fails over. It also gets the status of the secondary and can alert on unhealthy secondaries. If you have a secondary open read-only and want to monitor that activity, point…

Hi vrkvvj, here is a custom query that mines the DPA repository for the number of active sessions. Is this what you are looking for? Global Resource Metric Report - Custom Queries - Database Performance Analyzer - THWACK

Hi @"HrachyaH" , to make a DPA alert run at a specific time, you can update the time information in the DPA repository for this alert. Here is another Thwack.com article that should help: Daily Custom Alert - When will it run? - Forum - Database Performance Analyzer (DPA) - THWACK

Here is another example of finding queries that have had query execution anomalies: (+) Query Execution Count Anomaly - Custom Alerts - Database Performance Analyzer - THWACK (solarwinds.com)

Check out this custom query in the Content Exchange are for DPA: Top SQL by Any Dimension This one only uses a single dimension in the WHERE clause, but you can use this to create your own. And I see you did that and your query looks great as well!

A commit placeholder is telling you that there are waits when committing data. Look for other update/insert/delete/merge queries running at the same time as the commits, and that's what's causing it. Some solutions are commit less often (think of a loop updating 1000's of records and committing inside the loop for each…

They should be visible within DPA but it's possible they are running so quickly there is very little wait times for them so they are further down the list of SQL. However, if a report does not find them and you are searching for the proc name, then they are mostly likely not performance issues or anywhere close. You can…

You can create an Resource Metric alert and choose Round-trip time as the metric:

Unfortunately no, this is not supported.

At the bottom of the Database Instance Availability alert (actually any alert) you will see a Notification Policy dropdown. If you set the policy to "Notify when level changes" you will get notified when the instance is down, but then again when the instance comes back up and the alert goes back to normal status.

When you say DPA is not capturing SPs that run as part of a batch, what do you mean exactly? Do you have an example code and what DPA is not capturing?