Xenserver 7: RAID array monitoring
-
I have asked this question before, but I guess now I am asking it a different way...
From what I have read, it seems XS itself can not monitor a RAID array on a hardware controller. I have a new XS with a fresh RAID 10 array created in XS. RAID is disabled in the BIOS, so it should not be an issue. I created the array across four disks, formatted it, mounted it, and can see it in XC. I was successfully able to create a VM and copy a VM to the array with no issues. Now, how to I keep an eye on it? How do I get alerts? I see you can able alerts, but I dont see anything about RAID or HDD status, just thoughput.
-
So you were using FakeRAID on the motherboard and now you are using software raid?
-
@mazterjedi if you're using MDADM you can monitor the array with
watch cat /proc/mdstat -
@DustinB3403 said in Xenserver 7: RAID array monitoring:
watch cat /proc/mdstat
Yep, everything looks good...How do i know when something goes south without constantly watching it, though?
-
@mazterjedi said in Xenserver 7: RAID array monitoring:
@DustinB3403 said in Xenserver 7: RAID array monitoring:
watch cat /proc/mdstat
Yep, everything looks good...How do i know when something goes south without constantly watching it, though?
Where it says 4/4 and [UUUU] you might have 4/3 and [UU_U]
Also you can set email alerts for if a drive does go bad so that you are emailed about it.
-
This website is what I followed for my system.
-
@DustinB3403 said in Xenserver 7: RAID array monitoring:
@mazterjedi said in Xenserver 7: RAID array monitoring:
@DustinB3403 said in Xenserver 7: RAID array monitoring:
watch cat /proc/mdstat
Yep, everything looks good...How do i know when something goes south without constantly watching it, though?
Where it says 4/4 and [UUUU] you might have 4/3 and [UU_U]
Also you can set email alerts for if a drive does go bad so that you are emailed about it.
Which the email alert has worked, and told me that I had a bad drive in my array.
-
if you do need to replace a drive, you have to follow a specific sequence of steps.
This topic here outlines those steps.
https://mangolassi.it/topic/7772/replacing-a-failed-drive-in-md-raid-10
-
@DustinB3403
Personalities : [raid10]
md0 : active raid10 sdc1[1] sdd1[2] sdb1[0] sde1[3]
3906764800 blocks super 1.2 512K chunks 2 near-copies [4/4] [UUUU]
bitmap: 0/30 pages [0KB], 65536KB chunkunused devices: <none>
Ahh, I see. 4 U's is good lol.
So i checked out that website and looked at adding the email alert to the mdadm. But how? Im sorry, I am very new at XS/Linux/Cent and all this CLI is throwing me for a loop. -
Follow the steps on the website.
sudo nano /etc/mdadm.confadd at the bottom
MAILADDR [email protected]Test with
sudo mdadm --monitor --scan --test -1 -
Yours should look similar to mine.
-
This is how the test email will appear.
You'll have to setup an smtp email though
(trying to remember where that was...)See Below. -
Duh here is a guide I actually wrote...
