2025-02-11 Setup
Purpose: Repeat the test from 2025-02-05 methodology following modificatin of scripts to work on FreeBSD to see if they still work on Linux. Secondarily, test to see if the "corruption in 15 minutes" was a fluke or is typical.
2025-02-11 Plan
Refer to the previous iteration for specifics. This is just a repeat of that.
Time to complete first sync:
real 3468.47
user 42.80
sys 2188.07
root@orcus:~#
Second syncoid took 20s. All looks good. Kicking off in separate terminals.
thrash_stir.sh
thrash_syncoid.sh
manage_snaps.sh
Testing in full at 1525. top shows I/O wait at 2.5 t0 4.6%, idle over 72%. Test finished with corruyption at 1727. Not as good as 15 minutes previously but still reasonable.