The RAID array is all set up and working :+1: I had some trouble getting macOS Time Machine to behave, but figured it out eventually. Two drives -> RAID 1 array -> Encrypted APFS format -> two APFS volumes (one for each computer to use as a TM backup destination).
The initial backup, over wifi, took a few days 😅 But it did finish, and incremental backups are quick!
@ashfurrow My last raid build was raid 10. Not actually the tenth version but combines raid 0 and raid 1. 4 drives, twice as fast as one drive.
@ashfurrow Striping using raid 0 is great for speed but multiplies the chance of failure which is why raid 1/0 is recommended for the mirroring. However, The performance charts I've seen say raid 5 (stripe,stripe,CRC) is better for SQL server as it writes to the drives faster.
@angristan not to my understanding. RAID is support by the OS independent of the file system. but within the APFS system (which is mirrored with RAID 1) there are two APFS volumes, which are like partitions except both have access to as much free space as is available on the disk.
@ashfurrow For future reference, time machine backups run on a very low QoS priority to avoid hogging the CPU. You can temporarily increase the priority by running this command:
sudo sysctl debug.lowpri_throttle_enabled=0
Run again with enabled=1 to revert it to normal speeds later (this not only affects time machine so could have unknown side effects on other processes, probably nothing serious tho)
@brunoph cool! would be ran on the machine that’s backing up over the network, or the machine which is serving the shared drive? both, probably 😂
This Mastodon instance is for people interested in technology. Discussions aren't limited to technology, because tech folks shouldn't be limited to technology either!