Take the 2-minute tour ×
Server Fault is a question and answer site for professional system and network administrators. It's 100% free, no registration required.

I've searched high and low for a whitepaper on storing uploads and serving content (primarily images) via a Windows 2003 web farm, but to no avail. Our hosting facility has a SAN, which we have access to via a shared drive on the web servers, but in the past we've had difficulties making this work with our .NET apps.

Just wondering what everyone else does?

share|improve this question

migrated from stackoverflow.com Jan 11 '10 at 14:30

This question came from our site for professional and enthusiast programmers.

5 Answers 5

up vote 1 down vote accepted

How are you given access to the SAN? Are they just creating the volume on the SAN then giving you the target name etc to add to the iscsi connectors on Windows? You shouldn't really have any issues using a SAN between multiple servers as long as you setup how it's being accessed correctly.

What are you trying to do with shared storage between web servers? You may want to look into setting up a proxy which that web servers can use and then have the proxy use the shared storage.

share|improve this answer

If only one server needs access to the data, you can assign it directly so that the server sees it as a local drive (if redundancy is required, you can use Cluster Services). If multiple servers will need access, you will probably need a server attached to the storage to share it out on the network, unless the storage device has native Network-Attached Storage capability, in which case you should use that.

In Windows 2008 Cluster Services, there is an option to use a shared drive among multiple computers (possibly 2008 R2 only), but it's only supposed to be used for Hyper-V and is not recommended for regular data.

share|improve this answer

There's a volume created on the SAN, and shared via a Windows file server in the same domain as the web servers. We were then mapping a drive to the file server, however as "Zypher" pointed out I think we should've used UNC paths. However, I'm still not sure if this is the best solution. Any ideas Shadow?

share|improve this answer

A couple of things really - firsly permissions, and getting the remote drive mapped at server boot (when there's no user logged on), and secondly what happens if the remote drive becomes unavailable - how does it get remapped? Should we be using a mapped drive at all?

I've tried sharing the folder on the SAN as a virtual folder in IIS, and setting the permissions in there, but have had difficulties with permissions there too.

I'd love to comment on performance, but I've never actually managed to get it working well enough to benchmark it. I was just wondering how other people do it?

share|improve this answer
1  
Services should never be reliant on Mapped drives. This is why we have UNC paths. Use those for anything that isn't an interactive user application –  Zypher Jan 11 '10 at 22:52

What did the difficulties stem from? Failure of storage, error message, problem in the applications themselves, permissions, non-scaling, speed,...? Generally speaking a SAN to centralize storage as a back end for data should work fine. It depends on what trouble you were running into that would narrow down what your next course of action should be (what was the bottleneck, or was there a specific failure or error you were hitting?)

share|improve this answer

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.