Pick the best LUN size
This tip is excerpted from a discussion thread posted on the searchStorage Web site. The question posed was from user RickLR, who asked what issues he needs to be concerned with when determining the proper Logical Unit size for his virtual storage. Users ZiggyS and Alan McLachlan, who offer two different approaches to the solving this particular problem, joined the discussion. Basically, ZiggyS says smaller LUNs, but there's a downside. Alan says it's not quite as simple as all that.
What are the issues I need to be concerned about when determining a LUN size?
On our IBM Shark I can create one LUN over an entire raid set that shows up at the host computer as one disk, or, I can create several smaller LUNS and have several disks at the host computer. Which is better? I was told once that it was better to have multiple smaller LUNS than one big one since many smaller LUNS would allow more concurrent I/O's from the OS.
This would also apply to our EMC Symms with meta-volumes.
The advantage of many, smaller LUNs is that you can spread them over many array groups and use LVM to create you volume groups (concat or RAID-0). This will give you the best potential performance on the back-end. It is not the LUN size though. It is the ability to distribute the I/O across many array groups. On Shark,
The downside of many smaller LUNs is that you eat up internal addresses. If you use a LUN size of 2GB, then the maximum capacity of your Shark is 8TB (4096 x 2GB) no matter what size physical disk you use.
You said: "The advantage of many, smaller LUNs is that you can spread them over many array
groups and use LVM to create your volume groups (concat or RAID-0). This will give you the best
potential performance on the back-end." Maybe so, but you then create more load on the application
server. Whether using a larger LUN or using LVM to concatenate or stripe smaller LUN's is better
will depend on a whole bunch of factors related to the applications, server capacity, caching and
RAID performance on the subsystem, host OS file system management characteristics, etc. You don't
want to make your "back end" potentially better, only to find it doesn't get driven to that
potential because your host is slowing things down. I'm not saying you're wrong, just that it's not
necessarily that simple.
Of course, if you have to run the LVM for ANY reason, than it will add processing to all I/O's regardless so you might as well get the best you can from it. At the end of the day, server-side fs issues can subvert block I/O characteristics anyway, so a little extra overhead at the block device layer can become moot (which makes you right in those cases as well).
You can read the entire thread from which this tip is excerpted in the forum.
1. What are LUNs?
This EMC site has a great listing of storage terms and their associated definitions. It covers the gamut of terms, including: disk striping, LUNs, and SCSI. (Editor's note: Besides this listing from EMC, our own sister site -- Whatis.com -- is also worth a glance to help define just about any technical term. It will be a welcome addition to your list of favorites.)
2. Are there guidelines for creating metavolumes?
This reader posed an interesting question to SAN expert Christopher Poelker: "I was wondering what guidelines people are using for creating metavolumes. Does the space request need to be extremely large?" Chris discusses the importance of maintaining performance by spreading application load and the benefits of hardware-based LUNs to do the job.
3. What is the effect of having multiple HBAs in the same zone?
SAN expert Chris Poelker says the answer depends on two things: 1) The storage vendor equipment provides a means of LUN masking by World Wide Name in the storage arrays connected to that zone and 2) There is some sort of path filter driver on the host to enable fail over to visible LUNs.
4. What do I need to know about storage virtualization?
Virtualization. Everybody's talking about it. But, if you ask 20 vendors what it means, you may get 20 different answers. Especially when it comes to where it's located. These resources in our featured topic should help.
This was first published in November 2001