The Single Pane of Glass

Question

The Single Pane of Glass

Steve Jones - SSC Editor

SSC Guru

Points: 736255
More actions
October 4, 2018 at 8:13 pm

#364562

Comments posted to this topic are about the item The Single Pane of Glass

Viewing 8 posts - 1 through 7 (of 7 total)

You must be logged in to reply to this topic. Login to reply

call.copse SSCoach Points: 18053 More actions · Answer 1

That would be sweet Steve. Sadly there's stuff all over the place e.g. past double hop rdp servers and it would be tough to do that well right now. Excuses, excuses I know - we could certainly get a lot closer than now though. If I had time...

DinoRS SSCrazy Points: 2683 More actions · Answer 2

My current customer has an "IT-Dashboard" basically it shows if a certain application is working or not through various checks.

Unfortunately (as with most bigger organizations) there is no actual "Performance Dashboard" or something similar for DB Servers. I'm currently in the process of establishing the collection of such Information - you gotta start somewhere, right? I'm looking at basically wanting to have a mix of PowerBI (for Presentation) with SSIS Execution metrics (this is a BI Environment after all) paired with information like DB Growth Charts, Changes in Execution Plans over time etc. but it'll take a while (being the currently only Dev on this part) until I'll be able to present useful "IT decision making" metrics to let's say the CIO so he'll much more likely approve the addition of resources like CPU or RAM in a much shorter time.

Throwing Hardware at things that are already unusable isn't the challenge here, it's preventing things to get there.

MattF SSCrazy Points: 2285 More actions · Answer 3

We use several tools such as DLM Dashboard, SQL Change Automation, Azure boards and others to support our dev ops. For SQL Server monitoring we have developed dashboard reporting around SQL Data Collections and Reporting Services. This uses the CMS to generate a list of servers which the reports run against. The benefits of this approach are the stability of using the proprietary Microsoft MDW as a source, instead of a custom developed or third-party solution, and the associated costs and on-going maintenance for those alternatives.
This solution also works for our external DBA customers. Reports are emailed each morning, for current status and also trending. This way we don't have to remote in to their network to know how their system is performing.

I also like the look of the dashboards (insights) for performance monitoring in Azure Data Studio - hopefully these will become a bit more user-friendly as currently they are pretty unwieldy to setup.

MattF

richj-826679 Right there with Babe Points: 718 More actions · Answer 4

If there's a decent dashboard for SS, Win, Unix/Linux, Oracle, and Networking, I'd like to see it. Spotlight works well for us (smallish-to-medium business with a lean staff) for SS and Windows dashboard and alerting, but it doesn't do the rest. Oracle has it's own Enterprise Manager, but it doesn't monitor other things well. Zabbix for Unix, SolarWinds for LAN/WAN/Networking. I've worked with Nagios and have unfortunately been exposed to HP OpenView once upon a time.

SolarWinds and Nagios and others can monitor anything, but the setup/maintenance doesn't seem to make them a good fit (not to mention cost for some of them!). A better solution for me would be a NOC-like 4-6 block of monitors, each showing a dashboard for what they monitor best. I'm currently using Grafana as a front end for Zabbix and it just might be the SPoG for some, but we don't currently have anything that it could use for SS.

Fun!

Rich

Jeff Moden SSC Guru Points: 1004362 More actions · Answer 5

It certainly doesn't apply to everyone but I've found a lot of people that have such proverbial "Single Pane of Glass" tools either don't know how to use them to realize a problem is occurring or they don't know how to isolate the problems never mind actually fixing them. A great example of this is when people find certain wait stats that have gone off the scale. The first thing they do is blame the hardware instead of looking for poorly written code that has stacked up in the gullet of SQL Server because no machine could do otherwise. Such mistakes also lead to the erosion of the notion of DevOps because it turns into a blame game, correct or not.

--Jeff Moden

RBAR is pronounced "ree-bar" and is a "Modenism" for Row-By-Agonizing-Row.
First step towards the paradigm shift of writing Set Based code:
________Stop thinking about what you want to do to a ROW... think, instead, of what you want to do to a COLUMN.

Change is inevitable... Change for the better is not.

Helpful Links:
How to post code problems
How to Post Performance Problems
Create a Tally Function (fnTally)

Steve Jones - SSC Editor SSC Guru Points: 736255 More actions · Answer 6

the more complex your environment is, the less likely you'll have one screen to view it. However, you could have one screen that lets you know where things need investigation, and then you'd need to delve off into a more detailed view.

What I really wish is that more tools provided an API to let me pull certain data over, that would give me a high level view of all other tools, which I could then use to decide which area needs more investigation.

Eric M Russell SSC Guru Points: 125561 More actions · Answer 7

What we use is a combination of sql traces, pings, event logs, application APIs, and for some web applications: parsing http responses. All of it duct taped together with PowerShell and then aggregated into a database. From that central location, multiple user groups (DBAs, operations, executive management) can then create dashboards or PowerBI reports that allow them to focus on whatever facet of the organization they're interested in.

"Do not seek to follow in the footsteps of the wise. Instead, seek what they sought." - Matsuo Basho