For the past several weeks, after upgrading to 6.4.0 and then later to 6.4.1, my ReadyNas 312 has become nearly unusable; there is constant disk churn apparently due to btrfs-cleaner and btrfs-transaction that makes all other operations slow (if they succeed at all). The output of top, iotop, iostat, etc suggests that the system CPU is spending most of its time in iowait.
Sample output from iotop:
root@readynas312-1:# iotop -o -b -d 10
Total DISK READ: 642.96 K/s | Total DISK WRITE: 0.00 B/s
TID PRIO USER DISK READ DISK WRITE SWAPIN IO COMMAND
1939 be/4 root 1695.09 K/s 2.91 M/s 0.00 % 52.77 % [btrfs-cleaner]
Total DISK READ: 1451.80 K/s | Total DISK WRITE: 0.00 B/s
TID PRIO USER DISK READ DISK WRITE SWAPIN IO COMMAND
1939 be/4 root 1426.37 K/s 1849.12 K/s 0.00 % 92.83 % [btrfs-cleaner]
Total DISK READ: 944.06 K/s | Total DISK WRITE: 0.00 B/s
TID PRIO USER DISK READ DISK WRITE SWAPIN IO COMMAND
1939 be/4 root 943.27 K/s 1705.75 K/s 0.00 % 95.09 % [btrfs-cleaner]
I'm seeing consistant failures to mount AFP shares; when I go to the Shares tab in FrontView, they rarely appear there (the progress spinner runs for about 30 seconds before going away, and showing no shares). However, the shares appear from my client and are (sometimes, sometimes not) mountable.
Re: Constant btrfs disk churn - 6.4.1 nearly unusable
You may wish to open an online support case and attach your logs zip file to the case.
Do you use snapshots?
Which services and apps are you running?
How long have you left the system running for btrfs-cleaner to complete doing what it needs to do? In some cases after the upgrade to 6.4.x it can take quite a while for it to do what it needs to do.
