It is 8:52 in the morning. The first message lands before your coffee does: "Email is down." Then another: "Is it just me, or is the whole thing slow?" Then a third, from someone who never messages you, which is how you know it is real. You are the IT lead at Northwind Logistics, the one person everyone pings when something stops working, and right now three things have stopped working at once.
So you do the thing you always do. You open the M365 admin center in one tab. You check the AWS console in another. You search for whether anyone on the internet is complaining yet, because that is somehow faster than the official pages. 10 minutes go by and you still cannot say the one sentence everyone is waiting for: is this us, is this them, and do I need to do anything about it.
That 10 minutes is the cost of not having a single place to look. A service health status page exists to give you back those 10 minutes. One screen, one glance, and you know whether to roll up your sleeves or pour the coffee and wait. Here is how to read it.
What the tiles are showing you
When you sign in, the Service Health page shows a grid of tiles, one per platform Cloud Sentry watches for your organization. Each tile carries three pieces of information: the current status, when that status last changed, and a link out to the vendor's own status page if you want the source.
The grid is the point. You are not checking five vendor pages and stitching the answers together in your head. The platforms your team depends on, the M365 services and the AWS pieces behind your apps, sit in one view that updates on its own. The page renders with the most recent status on file, then the tiles poll for fresh data in the background, so the grid stays roughly live without you hitting reload.
The four colors, in plain terms
The tiles use a traffic-light pattern, with one extra color for the case nobody likes to admit happens.
- Green means operational. No known issues, and recent checks all passed. This is the color you want and the one you should mostly see.
- Yellow means degraded. The vendor is reporting a partial outage, a performance problem, or an in-progress maintenance window. Things mostly work, so expect slowness or the occasional error.
- Red means a major outage. The vendor is reporting that a core feature is unavailable to a large number of customers. This is not your laptop and not your config.
- Gray means we cannot reach the status source right now. Read it as "unknown," not as healthy and not as down. The thing you are watching might be fine; we just cannot confirm it this second.
The color that trips people up is gray, because instinct fills it in as good or bad. Resist that. Gray is a missing signal, and clicking into the tile usually tells you why.
Click in when you want the story
A color answers the first question. The second question, what is going on and for how long, lives one click deeper. Open any tile and you get a per-platform page with more detail, including the most recent incident timeline and a direct link to the vendor's status page.
That timeline is the part worth reading before you message anyone back. It is the difference between "email is down" and "M365 is reporting a partial outage affecting Exchange Online, vendor is investigating, started 8:30." One of those sentences calms a room. The other starts a second fire. The tile gives you the calm one.
When a tile is asking you to file
Here is the part that saves everyone time. When a platform shows yellow or red, you do not need to file a ticket to report the outage itself. Cloud Sentry is already watching that signal, which is the whole reason the tile turned color before your inbox did. Reporting a known outage just adds a ticket to a queue without moving anything forward.
File a request only in these cases:
- The outage is blocking specific work and you need a workaround to keep moving today.
- You are seeing a problem the tile is not reflecting, which is genuinely useful signal for us to have.
- You cannot tell whether what you are seeing is the outage or something separate on your side.
Notice what that list is about. Every case lives in the gap between the outage and your own work, which is the part a person can do something with. Vendor outages are weather, and you cannot file your way out of weather. Security and operations are not a screen you check; they are the judgment about what to do with what the screen shows, and that judgment is the work we run for you. The tile tells you the weather. We help you decide whether to bring an umbrella.
A status page does not fix the outage. It tells you whether the outage is yours to chase, and most mornings it is not.
So walk back to 8:52, three messages deep, no coffee. The version of that morning with a service health status page is shorter and quieter: you glance at the grid, M365 is yellow, the timeline says the vendor is on it, and you send one calm message; no five tabs, no guessing. The outage is the same. Your morning is not. The question worth asking is simple: when something breaks next, do you want to spend 10 minutes finding out it was never yours to fix?