Slack issues | Hacker News Reader

Slack issues

status.slack.com

102 pointsfloriangosse65 comments

Slack is definitely down even though their status page says otherwise.

Incentives matter. If there is no penalty for misrepresenting the current state of an outage, you are going to find that company driven status pages are going to be frequently inaccurate.

In fact, there's an inverse incentive at play here. Slack is on the hook for missed SLA, so any admission of issue is legally and financially culpable. This means they don't want to admit to an issue until it's impossible to deny there is one.

There is also an incentive to display the proper status, since if you show all green when the service is down, you look dumb.

We recently improved our status reporting because our old status system might not accurately indicate if only a part of our service went down, so it might show green even though (some) customers definitely saw very obvious downtime.

When the choice is between temporarily "looking dumb" and permanently loosing a bunch of money because of SLAs, companies like Salesforce usually make the latter choice.

Usually this is less being outright deceptive (saying you're up when you know you're down), and more wrapping the definition of "down" in legalese so that nothing short of a complete system outage of every aspect of your service for 100% of users counts as "down".

In those cases status pages are more likely to have a lovely spectrum of yellows and oranges to choose from, but a distinct absence of red, but very few companies will say that they're fully green when they know there's issues. Green when you're down is usually more of a symptom of process problems in how outages get reported or resolved (no documentation, manual processes for updating, long communication chains from support teams to engineers).

Nah the issue isn't looking dumb, the issue is looking like a liar. What good is an SLA if your reputation is that you're going to lie to avoid paying out?

This is Salesforce we are talking about. They topped Stack Overflow as the most dreaded technology so they bribed SO not to include them anymore. They don't care if they look like liars to technical people. Technical people are not the decisions makers in their target customer base.

> They topped Stack Overflow as the most dreaded technology so they bribed SO not to include them anymore.

Wait what? And SO agreed? Where can I learn about this?

I feel like customers treating you like a bunch of fucking idiots is not gonna be a great boost to your bottom line either. Maybe if you're in a market sector where reputation doesn't matter and you don't have any competition swooping in.

The people signing the contract with Slack is not looking at the status pages, generally.

The people signing contracts ONLY look at the status page, and say everything’s fine. They don’t bother verifying it signing in.

> There is also an incentive to display the proper status, since if you show all green when the service is down, you look dumb.

How many customers will think you look dumb? If you admit downtime, will the accurate reporting drive away more new customers, than the inaccurate reporting drives away due to you looking dumb?

"Looking dumb" is not a reasonable factor. Damage to your public image is, but that's divorced from looking dumb. It's entirely possible Slack is saving money by not admitting their downtimes, because it leaves a better image.

And yet, this generally doesn't happen.

Slack now has the outage listed:

https://status.slack.com/calendar

The online cynicism is something I've never understood. It assumes the world is filled first first-order simpletons & somehow only the poster rises above it.

People aren't dumb & companies aren't dumb. People will notice a status tracker isn't accurate. They will have reduced confidence in the company. Companies know people will notice & think this way, so they generally don't engage in this type of deception. I'm sure there are exceptions, but in general, status pages have tended to be accurate in my experience.

You can't measure "reduced confidence" like you can a dollars going out for not meeting an SLA, and what are you going to do in this case? Move to Microsoft Teams? Please.

Their status page is not a report of status, but rather a report of admission.

We noticed about 2 minutes of hard downtime (API and website) on OnlineOrNot: https://slack.onlineornot.com/

Another status page that sucks. Slack goes down, people start texting me about it, status page is green, HN informs me that Slack is actually down.

Next time I build a status page it will simply be a static HTML with green indicators and some random metrics, looks like it is the industry standard. :facepalm:

Progression of status pages, from experience at a large cloud provider...

Stage 1: Status is manually set. There may be various metrics around what requires an update, and there may be one or more layers of approval needed.

Problems: Delayed or missed updates. Customers complain that you're not being honest about outages.

Stage 2: Status is automatically set based on the outcome of some monitoring check or functional test.

Problems: Any issue with the system that performs the "up or not?" source of truth test can result in a status change regardless of whether an actual problem exists. "Override automatic status updates" becomes one of the first steps performed during incident response, turning this into "status is manually set, but with extra steps". Customers complain that you're not being honest about outages and latency still sucks.

Stage 3: Status is automatically set based on a consensus of results from tests run from multiple points scattered across the public internet.

Problems: You now have a network of remote nodes to maintain yourself or pay someone else to maintain. The more reliable you want this monitoring to be, the more you need to spend. The cost justification discussions in an enterprise get harder as that cost rises. Meanwhile, many customers continue to say you're not being honest because they can't tell the difference between a local issue and an actual outage. Some customers might notice better alignment between the status page and their experience, but they're content, so they have little motivation to reach out and thank you for the honesty.

Eventually, the monitoring service gets axed because we can just manually update the status page after all.

Stage 4: Status is manually set. There may be various metrics around what requires an update, and there may be one or more layers of approval needed.

Not saying this is a great outcome, but it is an outcome that is understandable given the parameters of the situation.