xfs
[Top] [All Lists]

Weird behaviour of mkfs.xfs

To: "xfs@xxxxxxxxxxx" <xfs@xxxxxxxxxxx>
Subject: Weird behaviour of mkfs.xfs
From: Emmanuel Florac <eflorac@xxxxxxxxxxxxxx>
Date: Sat, 26 Mar 2016 14:50:37 +0100
Delivered-to: xfs@xxxxxxxxxxx
Organization: Intellique
Hi all,
I'm testing NVMe storage, and it works fine so far. However here's
something bizarre: I've made a RAID-0 array of 2 NVMe drives with
mdadm, and mkfs.xfs /dev/md0 starts an endless, full speed write of the
array:

avg-cpu:  %user   %nice %system %iowait  %steal   %idle
           0,21    0,00    0,17   12,39    0,00   87,24

Device:         rrqm/s   wrqm/s     r/s     w/s    rMB/s    wMB/s avgrq-sz 
avgqu-sz   await r_await w_await  svctm  %util
sda               0,00     0,00    0,00    0,00     0,00     0,00     0,00     
0,00    0,00    0,00    0,00   0,00   0,00
nvme0n1           0,00     0,00    0,00 5152,33     0,00  1288,03   511,98  
1019,53  197,87    0,00  197,87   0,19 100,00
nvme1n1           0,00     0,00    0,00 5154,67     0,00  1288,67   512,00    
50,62    9,82    0,00    9,82   0,19 100,00
dm-0              0,00     0,00    0,00    0,00     0,00     0,00     0,00     
0,00    0,00    0,00    0,00   0,00   0,00
md0               0,00     0,00    0,00 5152,00     0,00  2576,00  1024,00     
0,00    0,00    0,00    0,00   0,00   0,00


I'm running a plain vanilla 4.4.6 kernel; the xfs_progs are only
slightly older, being version 4.2.

Nothing will stop the crazy mkfs.xfs, but  a hard reboot. Any idea?

-- 
------------------------------------------------------------------------
Emmanuel Florac     |   Direction technique
                    |   Intellique
                    |   <eflorac@xxxxxxxxxxxxxx>
                    |   +33 1 78 94 84 02
------------------------------------------------------------------------

<Prev in Thread] Current Thread [Next in Thread>