Are Time Machine backups completely fail proof? [closed]

During OS update, some people report that the update fails and turns the machine into a brick, forcing them to take it to Apple.

Fail proof: No matter what, you can revert back.

Can I trust Time Machine? There's no Apple service in this country.

2015 mbp 13" from 10.14 to latest stable, time machine backup files are on external device obviously


Solution 1:

+1 to what everyone is telling you about backups and failures in general. To address your question more directly, though: It is ALWAYS possible for an upgrade to brick your machine. The way this happens is during the firmware update sequence. A botched sequence and/or a buggy firmware image, either EFI or SMC, will render your machine instantly unbootable, and no amount of backups will help you. The ONLY way to make that machine boot again is to manually reflash the broken firmware, which requires special tools that you're very unlikely to have, and therefore must get a third party to help service your Mac.

The good news is that Apple goes to extensive lengths to prevent the above scenario, precisely for the reason that a bad firmware update could affect millions of customers and therefore also be catastrophically bad for Apple. That's why there are many firmware QA tests that have to get passed before any OS update is made available to the public. Testing comprises a significant percentage of the time that it takes Apple to release an update. The high risk of catastrophic failure is also why minor OS updates (as opposed to major OS upgrades) introduce few new features and are instead focused on bug fixes and polish: Making fewer firmware changes reduces risk.

In your case,

2015 mbp 13" from 10.14 to latest stable

you are attempting to upgrade a relatively old machine from Mojave to, presumably, Monterey. This is a +3 major version point jump. Such a procedure, when you can't afford a failure, is not recommended.

A lot of code at all layers of the stack changes between major upgrades. This constantly leads to situations where Apple has to write migration-specific code that will take data (including nonvolatile, boot-critical data) stored using yesteryear's paradigm and convert it to a newer format.

Apple's upgrade process is most heavily tested for the most common use case, which is a +1 major version upgrade. There is a lot less test coverage given to big leaps like the one that you are taking. The chances that a migration will run into an unhandled corner case are significantly higher when you are leapfrogging 3 major versions.

The safer approach (beyond having your data backed up in multiple locations) is to do major OS upgrades one-by-one (i.e, Mojave -> Catalina -> Big Sur -> Monterey, rather than Mojave -> Monterey). Yes, this takes longer, but it reduces the risk that you're going to end up having to erase & restore from your backups (at best) or have a system that no longer boots (at worst).