On Wed, 2019-11-27 at 14:03 +0000, Kay Diederichs wrote:
> Hi Vaheh,
>
> RAID on Linux comes in different flavours and levels; the flavours are
> software RAID (mdadm) and hardware RAID (dedicated RAID controller or
> motherboard), and the levels are RAID0 RAID1 RAID5 RAID6 RAID10 and a few
> others. These details influence what the user will notice when a disk goes
> bad. Without knowing what you have, it is difficult to help.
>
> As an example, by default in my lab we have the operating system on mdadm
> RAID1 which consists of two disks that mirror each other. If one of the disks
> fails, typically we only notice this when inspecting the system log files.
> Replacing the disk, and re-silvering the RAID1 is not trivial and requires
> some reading of material on the web.
>
> It sounds like you don't have this type of RAID1, or maybe there is some
> mis-configuration.
>
> good luck,
> Kay
>
Just to add a bit of advice based on long experience: Unless you really
do it every day, looking at system logs can be an unreliable way to see
when you have a bad disk. Assuming you have root access on your Linux
system, there is a tool called logwatch that can be set to look at your
logs for you and send you an email every day with selected excerpts.
So I add a script (I call it runmdadm) to root's /etc/cron.daily to
run mdadm every night and get the RAID info into the log. I use egrep
in the script to pick out the important output. Part of my script looks
like this (details changed):
#!/bin/sh
# Run mdadm to check status of RAID arrays
echo "mdadm /dev/md0" > /var/tmp/mdadm.log
/sbin/mdadm --detail /dev/md0 | egrep 'active|removed' >> \
/var/tmp/mdadm.log
Then you can follow the instructions for logwatch to add this new
/var/tmp/mdadm.log to the list of logs it looks at. (Or change the
above script to add the mdadm info to a log logwatch already looks at.)
If you have hardware RAID, the idea is the same, but you need to use
a command to the hardware controller instead of mdadm.
George Reeke (an old lurker)
########################################################################
To unsubscribe from the CCP4BB list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB&A=1
