Custom reports against alerts, queries and metrics
CONTEXT: We want to see the bigger picture arising from the data presented in SQL Monitor.
PROBLEM: Currently you can't pull together a report in the SQL Monitor UI, and the monitoring database is not particularly accessible. We need reports against:
- The metrics shown in the analysis graph
- The top queries and waits
- The alerts raised over time
- Server configuration and SLA
- Provide reporting scripts against the monitoring database
- Privide a BI interface against the monitoring database
- Priovide richer data-export functionality in the UI
- Provide a report designer in the UI (perhaps allowing for regular reports to be sent)
Some time ago I submitted a request 'Wait stats - add a functionality for graphical analysis and comparison of historical data', but it was merged into this request here.
I'm not sure it the "Reports" tab format is the optimal one for this.
Wait stats analysis and compassion are essential when troubleshooting performance issues.
I'll repeat the comment I made on the suggestion:
"In v4.0 wait stats analysis was added on the overview pages. I'd like to see that extended to the point where we'd be able to graphically analyse historical data about wait stats, similar as metric analysis on the Analysis tab. Baselines and comparisons included."
Do you guys (@RedGate) still feel this belongs in the "Reports" tab?
Already emailed these, but posting here as well:
- I suggest you add more periods (1 hour, 3 hours, 6 hours).
- Reports are now static. A switch to enable/disable auto refresh of the report would be nice – as in, if you want to use it as a dashboard on some big screen. Example: report for last X hours, but you want the charts to always show time up to the present moment. Report refresh frequency is debatable… perhaps make even that a choice, per report (every 10 seconds, every X minutes…).
I have some feature requests for the new reports function in sqlmonitor (188.8.131.5203).
- copy reports: usecase – is define a report for a server + copy this report for other servers
- view time: please make this filed as free definable – usecase – create reports for an whole month, not last 27 days
- disk avg. read/write: I want to choose the scaling by my self – usecase – to ignore a spike
- disk avg. read/write: please specify the term TimeSpan on the ordinate with seconds, millisecond
It will be great if there was instead of top 10 worst performing query top 50 for example and choose a query/SP and check what was the difference today compared to 2 months back in the time in term of duration and some wait stats. + the scheduled report to be sent via email in PDF or HTML format. As well apart from Availability Groups no other DR metrics are available custom or build in (Replication,Mirroring,LogShipping)
Morten Nilsen commented
Is there a way to embed one of these report tiles in a custom dashboard?
I currently load MRTG Weathermap images and splunk reports as iframes into a simple webpage that I have open on an overhead screen to monitor our systems. I would love to get SQLMonitor up on this screen as well.
This is something we are currently developing. We'd like to understand a bit more about the requirements and get feedback on some possible solutions - anyone that would like to contribute, please email email@example.com
I submitted this as a request about a year ago.
Prasanth Matikana commented
Generate scheduled reports on every added instance.
For example: A weekly health report or Weekly long running query report or monthly Database growth report etc.
Garry Armour commented
Hi, I have a quick comment to make on this request.
I have a situation where I need to demonstrate an improvement or drop in number of deadlocks, blocked processes, long running, etc... over time
So, for example, an increase in CPU count on an SQL server has had a marked improvement on the performance but I cannot export any alerts from SQL Monitor to show this.
Ideally, an ability to export the alerts based on the current filter would be very useful. csv, excel, whatever.
Morten Nilsen commented
For building a global monitoring dashboard, I would like the ability to set up an analysis graph and embed it as an image from a different webpage.
Some way to embed basic stats/alert counters and the like would also be handy.
I would love to see a enhancement for sending out an automated mailer daily/customized schedule for giving a report for each and every server with common important metrics to analyze the server health on email without need to login into RedGate.
When running trace through redgate can you add the option to export the profile so when I click on a spid I can export the trace as you can copy the whole content.
Dag Calafell commented
I'd like a KPIs area. For example:
Number of Occurrences above 60% CPU: 193
Average Duration of an Occurrence over 60% CPU: 68 Seconds
Number of Occurrences above 60% CPU which lasted longer than 3 minutes: 10
Total time spent above 60% CPU: 221 Minutes
Longest occurrence above 60% CPU: 30 Minutes
Same for page splits, PLE, et cetera
This would be a GREAT feature to add. My boss wants to see how our SQL Servers are performing CPU and disk utilization wise on a weekly basis.
Thaks for raising this as a new suggestion, we'll keep a close eye on this.
In v4.0 wait stats analysis was added on the overview pages. I'd like to see that extended to the point where we'd be able to graphically analyse historical data about wait stats, similar as metric analysis on the Analysis tab. Baselines and comparisons included.
Yes, it would be great to have the ability to export a single alert from the alert details page. It would also be great to be able to export the whole ball of wax...
Additional features for LRQ Analysis
Thanks for your suggestion, I've added this to our enhancement tracking
When receiving email alerts for slow running queries, it would be valuable (more and more so) to have an HTML page with an Export to XLS option, showing the alerts received via email and any/all associated metrics with the slow running queries.