I just redid my NAS, switching from ZFS to LVM raid. This is a new raid with 3x8TB, each physical volume is a LUKS volume.
lvs -a -o+devices
shows me:
LV VG Attr LSize Pool Origin Data% Meta% Move Log Cpy%Sync Convert Devices
[lvol0_pmspare] vgraid ewi------- 2.00g /dev/mapper/cache_data(359138)
lvraid vgraid rwi-aor--- 14.55t 90.92 lvraid_rimage_0(0),lvraid_rimage_1(0),lvraid_rimage_2(0)
lvraid_cache_data vgraid Cwi---C--- 1.37t lvraid_cache_data_cdata(0)
[lvraid_cache_data_cdata] vgraid Cwi------- 1.37t /dev/mapper/cache_data(0)
[lvraid_cache_data_cmeta] vgraid ewi------- 2.00g /dev/mapper/cache_meta(0)
[lvraid_rimage_0] vgraid Iwi-aor--- <7.28t /dev/mapper/sda1_crypt(1)
[lvraid_rimage_1] vgraid Iwi-aor--- <7.28t /dev/mapper/sdb1_crypt(1)
[lvraid_rimage_2] vgraid Iwi-aor--- <7.28t /dev/mapper/sdc1_crypt(1)
[lvraid_rmeta_0] vgraid ewi-aor--- 4.00m /dev/mapper/sda1_crypt(0)
[lvraid_rmeta_1] vgraid ewi-aor--- 4.00m /dev/mapper/sdb1_crypt(0)
[lvraid_rmeta_2] vgraid ewi-aor--- 4.00m /dev/mapper/sdc1_crypt(0)
while restoring my backups, all kinds of alarms went off, specifically the load going through the roof. even if I stop everything that's running on the machine (it has a bunch of docker containers), it stays at ~5-7.
iotop tells me that the process with the most IO is dmcrypt_write ... so i am assuming that this is because the raid is syncing all blocks, and that it should be over when the "Cpy%ync" colums hits 100%.
Is this correct?
.rm
