Suggestions
Welcome to the Redgate Monitor feature suggestions list. Tell us how we can improve Redgate Monitor by voting on these suggestions or by submitting your own ideas.
This is the easiest way to make an impact on Redgate Monitor’s roadmap. Check out the completed suggestions to see how your ideas have influenced us in the past. Please note that this list is one of multiple sources that drive our development, meaning that we will not always implement the highest-ranking suggestions.
To give your idea the greatest possible impact, please follow these guidelines:
- Describe the goal you’re trying to accomplish, rather than a specific solution to get there
- If you have a favoured solution, add it as a comment to your suggestion
- Give us as much context as possible
- Only make one suggestion per post. If you have multiple ideas, submit them in separate posts
- Comment and vote on existing posts to develop the ideas — the Redgate Monitor team will often stay quiet initially about new suggestions to see what you have to say!
If you have any questions visit the Redgate Monitor forum.
688 results found
-
Allow trace to be set at the individual alert level and severity
SQL Trace can highly impact the performance of the server. I can't benefit from trace right now as it's set on any alert to launch a trace.
Other monitoring tools, for instance Sql Sentry, allow the individual alert to be configured with settings for trace.
For example: A long running job might benefit from a trace, while a "dbcc checkdb " hasn't been run in a long time should not trigger a trace. This increases storage and server load for no benefit.
Please look into allowing traces to be configured based on the alert AND the severity level defined on…
6 votes -
Show a log of query activity
CONTEXT: When investigating a performance issue, I want to know what has been running
PROBLEM: SQL Monitor has rich query information on the overview screens, the performance diagnostics panels, and the query alerts, and the query trace in alert detail reports, but it's difficult to tie that information together
EXAMPLE SOLUTIONS:
- Implement a query log
- Show query information on the analysis graph45 votes -
Make comments apply to alert types, rather than individual incidents
CONTEXT: We use comments to document the root cause of alerts we encounter
PROBLEM: When alerts repeat, recovering the comments is a pain - they only apply to individual occurences, so don't get surfaced when they issues re-occur.
EXAMPLE SOLUIONS:
- Show historical comments in alerts
- When creating alerts, allow us to choose whether this applies to the whole alert typeADDITIONAL CONSIDERATIONS:
Comments should also show when configuring alerts32 votes -
Control Alerts Using Metadata
CONTEXT: We have a maintenance and backup solution which uses SQL Server as a backend to determine what databases to perform the tasks against, and at what time.
PROBLEM: Some databases do not have maintenance tasks running against them, but still have alerts which appear in SQL Monitor for them (e.g. Index Fragmentation). I don't want to manually exclude these from SQL Monitor, as the number of databases fluctuates, particularly on our testing and development environments.
EXAMPLE SOLUTIONS:
- Provide a means of easily enabling/disabling/changing alerts dynamically using a stored procedure.1 vote
Daniel Rothig
responded
For your usecase I recommend modifying the maintenance window metadata instead:
UPDATE s SET MW_IsEnabled = 1, —or 0 to enable alerts again MW_Start=0, —midnight MW_Duration=864000000000, —24 hours MW_Monday = 1, MW_Tuesday =1, MW_Wednesday=1, MW_Thursday=1, MW_Friday=1, MW_Saturday=1, MW_Sunday=1 FROM [settings].[SqlServers] s INNER JOIN [settings].[Clusters] c ON c.Id = s.ParentId WHERE c.Name = ‘hostname’ AND s.Name = ’’ —Set to SQL Server name or leave empty for unnamed instancesA restart of the base monitor service will be required after the change
-
Baseline machine metrics across cluster failovers
CONTEXT: Using the Analysis graph for historical performance analysis of machine metrics on a cluster that failed over at least once
PROBLEM: Machine metrics are collected per node. Baselines of only one node can be shown at a time, and so no baseline view of performance on the "active" component can be synthesised.
EXAMPLE SOLUTIONS:
• Enable multiple simultaneous baselines
• Implement synthetic (bimodal) baselines across cluster nodes
• Provide synthetic metrics for entites on the "active" component of a clusterWhat do you think about these solutions? Got another idea? Let us know in the comments!
18 votes -
Drop markers on the graphs and add annotations
While working on performance tweaks it is customary to consult the analysis reports on a regular basis to see if our changes moved the needle. More often than not the tweaks are released one small batch at a time and it'd be nice to drop markers and custom annotations directly on the charts. I'd like to use that feature to add short phrases describing what was changed that led to the improvement, or degradation for that matter, of performance.
This would be by server (cluster/node) and by metric with the same retention period as the base data for that chart.
…
61 votes -
Simpler Upgrade \ Patch Process
The upgrade process is a bit long winded. It would be great if the system could be upgraded or patch with a single click. Ideally within SQL Monitor but if not, downloading the file and just selecting upgrade would be great.
70 votesDaniel Rothig
responded
Thanks Phil – we are considering reworking the upgrade process. If this suggestions gains more traction, we’ll prioritise it higher
-
Analysis parsing of dates and multiple baselines using a text input search
Allow the analysis and other date filters to provide the option to paste and try parsing the date based on matching common patterns (maybe even allowing user to define options in config file).
For example, the standard SQL datetime2(0) of yyyy-MM-dd HH:mm:ss pasted would convert to the date it parses.
Maybe even consider taking a page from the freaking great design Simple Bank has started applying which is a text based filter applying friendly syntax. I have found this incredibly useful.
They let you type in the search filter
"last 3 months"
"last week"
"date:1/1/2015-1/15/2015"
"date:1/1/15-1/15/15"and it immediate updates…
3 votes -
Use extended events instead of trace when trace initialized due to alert
I have seen a related idea https://sqlmonitor.uservoice.com/forums/91743-suggestions/suggestions/5717202-use-extended-events-for-deadlock-tracing-instead
However, I'm expanding this. If the SQL version is high enough the trace of current activity should be using extended events instead of a trace, this would reduce the total impact to the server performance while also expanding the available information.
98 votesDaniel Rothig
responded
Thanks for the suggestion – we’ll monitor how popular it gets
-
comment markdown
CONTEXT: I use comments on alerts raised by SQL Monitor to document root cause of the triggered condition and resolution action.
PROBLEM: Comments only support plain-text. When creating/reviewing comments, it's often useful to add links for more detailed descriptions of the cause of the alerts, and links to ticketing systems for resolution steps. Simplified bulleting/numbering and bold/italicized text would also be helpful for breaking up monotonous text.
EXAMPLE SOLUTIONS: Add Markdown (http://daringfireball.net/projects/markdown/) support to addition/review of comments. A couple implementations for examples of usage:
- StackOverflow (http://stackoverflow.com/editing-help)
- GitHub (https://help.github.com/articles/github-flavored-markdown/)
4 votes -
Install SQL Monitor without needing to provision a (large) SQL Server database
CONTEXT: One challenge with installing SQL Monitor is that we need to provide a connection to a SQL Server database.
PROBLEM: This can be costly or complicated. It would be better if SQL Monitor had an alternative backend.
EXAMPLE SOLUTIONS:
• Offer a hosted data store
• Automatically deploy a free backend alternative such as mongoDB or PostgreSQL
• Reduce data collection so that SQL Express can be used for small deployments
• Provide greater granularity over data retention policiesWhat do you think about these solutions? Got another idea? Let us know in the comments!
153 votes
Adam
responded
We’ve made some improvements in this area but will leave it open so you can continue to give us your suggestions
-
Make alert configuration levels more discoverable
CONTEXT: In SQL Monitor, you can configure alert behaviour down to the object level (such as the Database or the Job level). This is useful e.g. when we have specific databases with different backup policies to our global configuration, since we can specifiy "special rules" for individual databases.
PROBLEM: It is counterintuitive how to do so, and many people don't spot the left-hand drilldown navigation - this makes you think that you can only configure on the global and the server level! Also, when we have created an override, it's easy to forget so causes surprise when we change the…
172 votescurrently under investigation
-
Select and store a static baseline for future reference
CONTEXT: When comparing server performance over longer periods of time or across deployments, the baseline functionality in SQL Monitor's analysis section is useful.
PROBLEM: SQL Monitor creates a dynamic baseline based on previous data. While this baseline is a realistic idea of what's "normal" it requires us to keep a lot of data in SQL Monitor
EXAMPLE SOLUTIONS:
• Aggregate data in SQL Monitor so that it can be retained for longer without using much space
• Allow us to "freeze" a baseline so that the data doesn't get deleted.What do you think about these solutions? Got another idea?…
41 votes -
Develop SQL Server Agent Job Status Alert
Add a new SQL Server Agent alert to monitor the status of a job. The ability to alert when a job is running longer than a specified time or has not run within a specified time. The current alert will only notify you after the job has stopped. This does not help when a job is hung.
12 votes -
allow analysis graphs to update with realtime data without having to manually refresh them
We have SQLmonitor analysis graphs running on wallboards to show us data that is useful to us, but we continually need to click the button to "jump to present", as the timeline does not update as more data is collected to show the most recent history. We want the graph to move automatically to match the current time of day.
Maybe a "play" button that would allow thye display to automatically refresh the graph along the timeline and a pause button should you want to drill deeper into a certain time period when investigating issues. Having the graph update automatically…19 votes -
In the Analysis tab, it would be better to have an option to change the font from grey (very hard to see) to black/dark blue or something
In the Analysis tab, it would be better to have an option to change the font from grey (very hard to see) to black/dark blue or something
6 votes -
More options for Custom Metric Frequency
Would it be possible to have a custom option for the Custom metric gathering, as it stands the data can be collected at intervals of 1,5,10,30,60,1440 minutes
it would be nice to be able to select a custom value and values less than a minute, the standard metrics that are gathered appear to be gathered every 20 seconds which isn't an option here
Thanks for your time and concideration
Ste9 votesDaniel Rothig
responded
Note that you can currently hack an arbitrary “seconds” value into the SQL Monitor database – but changing the Custom metric configuration via the UI will reset the collection interval to one of the default values.
-
Custom metric: alert when value changed
I want to be alerted if someone creates or deletes a database. The metric itself is easy (select count(*) from sys.databases - executed for the master DB on all servers).
But in the alert setting I can only define that it should trigger if the new result is abover or below a set value but not for value changed (compared eigther to the previous value or the "baseline").
So I have to create two metrics / alerts for DB drop (below) and create (above) AND have to define / modify the number of databases as alert threshold for each server…
19 votesDaniel Rothig
responded
-
Search through (closed) alerts for a string.
It would be great if we could search closed alerts for a string. For example I want to see all the alerts that have included a particular view in the query text.
(Also include comments in the search)
69 votesDaniel Rothig
responded
We’re looking into this – no ETA at the moment
-
Allow Fractional Alert Thresholds.
When you use the "Calculated rate of change" functionality on a custom metric, the results come back in fractional values and you should be able to alert on these values. In my case, I want to monitor number of orders for the last 10 minutes. There are always relatively few orders at night and many during the day, but if order flow drops off drastically in a short period of time, I want to know about it.
Unfortunately, I can't just alert when the rate of change drops by 25% (-.25) because alerts can only happen on integer values. 100%…
4 votesDaniel Rothig
responded
A current workaround is to multiply the result by a factor, eg. 100 to make the result a percentage.
- Don't see your idea?