We now perform a "chkdsk c: /scan" daily (sometimes more than one a day) on all client systems using OneDrive. We are suffering chronic file system corruption (reparse errors) on those systems, which results in many strange problems for our users. To make matters worse, when the "chkdsk c: /scan" detects a reparse error, it appears to have a bug with its online repair code, so rather than fixing the problem right there and then, chkdsk instead simply blows up! At least chkdsk marks the drive's file system as dirty (that is, the file system needs to be structurally checked at the next boot) before it crashed. The check it performs at the next boot is a "/f" type, so it typically takes less than 5 minutes to complete (as opposed to the much more time consuming "/r" type); however, it does mean making the user reboot in the middle of their day to fix the corruption.
As we can no longer count on healthy file systems for our business clients that make use of Microsoft OneDrive, we are transitioning off of the Microsoft 365 platform--and Windows--entirely. An operating system and cloud service that can't maintain the integrity of client data is absolutely useless and of no value. In fact, persisting with its use when you are aware that it is chronically corrupting data is irresponsible for our own data, and our duty to our client's data.
Microsoft seems to have made no attempt to address this problem in any meaningful way. Instead, it just keeps advising users to do a "chkdsk" to fix it each time it occurs. However, because the symptoms a damaged file system cause are so varied, our users don't know when to do the "chkdsk". Instead, they suffer with bizarre problems--sometimes for days! Microsoft's lack of action to correct the cause of the chronic reparse errors for so long just makes clear how little attention it pays to small business and other users of Microsoft 365 products.
I believe this marks the final chapter in Microsoft's role as a partner to small businesses. It's time to move on to greener pastures.