Cloud · Business continuity

Outlook was down for a day. Our Mailcow client did not notice

On 12 and 13 November 2026, Microsoft's Exchange Online was unavailable worldwide for about 24 hours. In Malaysia that was a whole working Friday. What happened, why our client on self-hosted email kept working, and what every business should take from it, whichever email it uses.

Pikabyte Engineering 4 min read

~24 h

Exchange Online mailboxes were unavailable worldwide on 12 to 13 November 2026

On the evening of 12 November 2026, UTC, mailboxes on Microsoft’s Exchange Online stopped responding. Outlook on the web, the desktop app and phones all failed. Microsoft opened an incident, then widened it to every Microsoft 365 service that depends on Exchange: Teams calendars, Copilot’s access to mail, parts of the admin centre.

Reports on Downdetector peaked at close to 140,000. Service returned roughly 24 hours later. Accounts of the exact window differ, and some describe degraded service for longer. Early reports pointed to a failure in the gateway layer that sits in front of mailboxes. Microsoft’s final root-cause report had not been published when we wrote this.

What it looked like from Malaysia

The timing was unkind. The first mailbox errors came at about 16:20 UTC on Thursday. That is twenty past midnight on Friday in Kuala Lumpur.

So Malaysian businesses on Microsoft 365 arrived at work on Friday 13 November to no email, and went home without it. Quotations did not go out. Purchase orders did not arrive. Anyone waiting on a customer’s reply did not know whether it had been sent. A whole trading day, with the weekend behind it.

Why our client kept working

We host email for a manufacturing client on Mailcow, an open-source mail platform, in a cloud account the client owns. Their mail does not pass through Microsoft’s service at all. It arrives at their own server and is read from their own server.

On that Friday their staff sent and received mail as usual. Business carried on. For them the outage was something they read about, not something they lived through.

That is not because our setup is cleverer than Microsoft’s. It is because it is separate.

The real lesson is concentration

It would be easy, and wrong, to conclude that self-hosting is more reliable than Microsoft. Microsoft runs one of the best-engineered services in the world, with resources no small company can match. A self-hosted server can fail too.

The difference is what fails together.

When a service used by a large share of the world’s businesses has a bad day, all of them have it at once. Your email is down, your customer’s is down, your supplier’s is down, and the support line you would call is overwhelmed. Nothing you do can fix it. You wait.

When your own server has a problem, it is yours alone. Your customers’ mail still works. And there is a person you can call who can act: restart it, restore it, move it.

A self-hosted server does not remove risk. It gives you a different risk, one that is smaller in scope and inside your control. For some businesses that is a better trade. For others it is not.

What self-hosting demands in return

We have written about the case for owning email infrastructure, and the costs have not changed.

  • Someone must maintain it. Updates, certificates, disk space, monitoring.
  • Deliverability is your job. Authentication records must be right, or your mail lands in spam.
  • Backups must be tested. A restore you have never tried is a hope.
  • The account must be secured. A mail server is a target, and the cloud account under it needs proper guardrails.

A neglected mail server is far worse than any vendor’s outage. Self-hosting only works with a competent owner, in-house or under a support agreement.

If you stay on Microsoft 365 or Google Workspace

Most businesses will, and that is a reasonable decision. But decide it knowingly, and plan for the next outage. There will be one.

1. Have a second channel that does not depend on email. A WhatsApp Business number or a phone line that customers already know. On the day, it is the only thing that works.

2. Keep key contacts somewhere else. If your customer list lives only in the mailbox, you cannot reach anyone when the mailbox is down.

3. Know where your documents are. If quotations and contracts exist only as attachments, a mail outage is also a document outage.

4. Decide who communicates, and how. One person posts a notice on the website and messages key customers. Decide this before you need it.

5. Ask what else stops. Many systems send through the same service: invoices from accounting, alerts from operations, password resets. List them.

6. Consider a standby. Larger firms keep a separate continuity mail service for exactly this. It costs little next to a lost day.

A question for the board

Email is the one system every business assumes will be there. Friday 13 November showed that even the largest provider can be gone for a full day.

So ask the question plainly: if our email stopped for 24 hours tomorrow, what would we do in the first hour? If nobody has an answer, that is the work to do, whoever hosts your mail.

Sources

Written by

Pikabyte Engineering

Pikabyte builds custom business systems, applied AI and secure cloud infrastructure from Malaysia. Talk to us.

The work behind this article

Case study · Manufacturing · Infrastructure

Full ownership of corporate email: the data, the policies and the bill.

We moved a manufacturer from a vendor-managed email subscription to a self-hosted platform in an AWS account the company owns, without interrupting mail flow.

100%
of mail data in an account the client owns
0
per-mailbox licence fees
Read the case study Soon Chuan Hong Hardware Sdn. Bhd. · Malaysia

Start a conversation

Want this applied to your business?

Tell us what is slowing the business down. You will talk to an engineer, not a salesperson, and get a straight answer on whether software is the right fix.