KDE Almost Lost All Of Their Git Repositories

[phoronix]

Phoronix: KDE Almost Lost All Of Their Git Repositories

There was almost "The Great KDE Disaster Of 2013" when the KDE project almost lost all of their 1,500+ Git repositories...

http://www.phoronix.com/vr.php?view=MTMzNTc

[chuckula]

Phoronix: KDE Almost Lost All Of Their Git Repositories

There was almost "The Great KDE Disaster Of 2013" when the KDE project almost lost all of their 1,500+ Git repositories...

http://www.phoronix.com/vr.php?view=MTMzNTc

funkSTAR was THIS CLOSE to getting everything he ever wanted!

Actually, while it would have been a huge PITA, there are enough out-of-repository copies of the source floating around so that I'm sure they could have rebuilt the repos to at least the last-released state of the code base. Still not good, but not lost forever either.

[birdie]

Yeah, Linux and Open Source are shaky.

In fact I had an ext4 corruption on a partition I mount RO daily and remount RW maybe once a week to write a file or two.

[freedam]

Go suck horse cocks with your mother, birdie!

DO NOT FEED THE TROLL PLEASE

[ryao]

This is the silent corruption that ZFS was created to prevent. Unfortunately, KDE's servers were not using it.

[birdie]

This is the silent corruption that ZFS was created to prevent. Unfortunately, KDE's servers were not using it.

The real problem KDE sysadms overlooked is that mirroring is not backing up. ZFS wouldn't have saved them if their HDDs totally failed - ZFS is good at spotting hardware errors - that's true. But ZFS is not a magic bullet.

[ryao]

The real problem KDE sysadms overlooked is that mirroring is not backing up. ZFS wouldn't have saved them if their HDDs totally failed - ZFS is good at spotting hardware errors - that's true. But ZFS is not a magic bullet.

It is possible that the backups themselves would have been corrupted had the administrators been doing proper backups and these issues predated the fsck. In this situation, silent corruption caused a problem occurred because there was no way to recover after its effects had been felt.

ZFS is not a replacement for proper backups, but it makes doing them easier. End-to-end checksumming and self-healing would have prevented this mess before it happened had ZFS been deployed.

[Ericg]

Phoronix: KDE Almost Lost All Of Their Git Repositories

There was almost "The Great KDE Disaster Of 2013" when the KDE project almost lost all of their 1,500+ Git repositories...

http://www.phoronix.com/vr.php?view=MTMzNTc

DO NOT COMMENT ON THIS THREAD UNLESS YOU READ BOTH THE BLOG POST AND THE FOLLOWUP. A lot of questions get answered in both that michael skimmed over or didn't mention.

[birdie]

DO NOT COMMENT ON THIS THREAD UNLESS YOU READ BOTH THE BLOG POST AND THE FOLLOWUP. A lot of questions get answered in both that michael skimmed over or didn't mention.

I've read both - they didn't have proper backups. Period.

[ryao]

I've read both - they didn't have proper backups. Period.

Unfortunately, proper backups would not have helped as much as one would think without a way to know whether or not the data was bad prior to doing them. It is not clear when the repositories became corrupted, although the blog post suggests that fsck was the trigger.