Kevin H. Hobbs wrote:
We recently set up a fancy file server running Fedora core 6 with a
3ware 9550SX raid controller. As far as I can tell the correct line
in /etc/smartd.conf to monitor the 5th disk looks like this:
/dev/twa0 -d 3ware,4 -H -m root
The device /dev/twa0 was not automatically created, but only showed up
when I ran tw_cli from 3ware. Once the device exists I can run
smartctl :
[kevin@backup ~]$ sudo /usr/sbin/smartctl /dev/twa0 -d 3ware,4 -H
smartctl version 5.36 [x86_64-redhat-linux-gnu] Copyright (C) 2002-6 Bruce Allen
Home page is http://smartmontools.sourceforge.net/
=== START OF READ SMART DATA SECTION ===
SMART overall-health self-assessment test result: PASSED
But when start smartd I get a SELinux message for each disk, and smartd
cannot access the disk. This is what I see in /var/log/messages :
Feb 28 13:42:27 backup kernel: audit(1172688147.177:55): avc: denied { getattr } for pid=3099 comm="smartd" name="twa0" dev=tmpfs
ino=111017 scontext=user_u:system_r:fsdaemon_t:s0 tcontext=user_u:object_r:device_t:s0 tclass=chr_file
Feb 28 13:42:27 backup smartd[3099]: Device: /dev/twa0 [3ware_disk_04], File exists, open() failed
Feb 28 13:42:27 backup smartd[3099]: Unable to register ATA device /dev/twa0 [3ware_disk_04] at line 7 of file /etc/smartd.conf
Feb 28 13:42:27 backup smartd[3099]: Device /dev/twa0 [3ware_disk_04] not available
How do I have the system create the device for the controller
automatically on startup with the correct security context for the smart
daemon?
semanage fcontext -a -t fixed_device_t -f "-c" "/dev/twa[0-9]*"
Will load the file context and then you need to restart udev and it
should get labeled correctly.
