Software-based VM-centric and flash-friendly VM storage + free version
Moderators: anton (staff), art (staff), Max (staff), Anatoly (staff)
-
robnicholson
- Posts: 359
- Joined: Thu Apr 14, 2011 3:12 pm
Mon Aug 04, 2014 11:00 am
This isn't really a StarWind question but I'm guessing some of you will be able to point me in the right direction.
Our current v6 production build has both what we call primary and secondary storage. Primary storage is on faster 15k SAS drives but due to 600GB limit, is more precious. Secondary storage is on 7k SATA drives that is supposed to be for less frequently used storage.
However, I happened to look at throughout on the two targets on StarWind over a 24 hour period and 23GB had been read from secondary storage compared to 15GB from primary storage. That at first glance looks the wrong way around! Sure, not a very scientific check so we raised a job to investigate further.
Starwind itself doesn't appear to give much information/reporting here aside from last hour or last 24 hours. Can't check back over a week - that's one for the wishlist. Maybe it's already in v8. More granular reporting would be good as well around iSCSI connections - that would allow us to determine which Hyper-V hosts were generating the load.
But I suspect we actually need to monitor VMDX usage on the Hyper-V cluster. So hence this post - any ideas how to go about this? Third party tool? We do actually have SCOM licenses but we're only just rolling out that system.
Cheers, Rob.
-
robnicholson
- Posts: 359
- Joined: Thu Apr 14, 2011 3:12 pm
Mon Aug 04, 2014 11:13 am
PS. the reason we were looking at throughput was we were pondering over what size of L2 SSD cache we might need. The D series from Intel look really cool:
http://www.intel.co.uk/content/www/uk/e ... s-ssd.html
But the price of the 1TB+ options are a little expensive so we wondered if we actually needed that much. Real rough extrapolation of the throughput figures from the SAN show 54GB read over 24 hours and 9.2GB written (sounds about right). So a 400GB SSD would easily have held all the data read during a day.
Yes, this isn't very scientific as 54GB read total seems very low but then again, maybe not. But hence the reason for more accurate monitoring of disk reads & writes.
-
Anatoly (staff)
- Staff
- Posts: 1675
- Joined: Tue Mar 01, 2011 8:28 am
-
Contact:
Sat Aug 09, 2014 12:15 pm
Can't check back over a week - that's one for the wishlist.
That is in the roadmap already
But I suspect we actually need to monitor VMDX usage on the Hyper-V cluster. So hence this post - any ideas how to go about this? Third party tool?
Why not? Veeam and 5nine got something for free.
Best regards,
Anatoly Vilchinsky
Global Engineering and Support Manager
www.starwind.com
av@starwind.com
-
Anatoly (staff)
- Staff
- Posts: 1675
- Joined: Tue Mar 01, 2011 8:28 am
-
Contact:
Thu Aug 14, 2014 8:53 am
Let me know if you`ll need anything.
Best regards,
Anatoly Vilchinsky
Global Engineering and Support Manager
www.starwind.com
av@starwind.com