Sign in to follow this  
Followers 0
Brian

SAN/NAS
Starboard Storage to Cease Operations Discussion

4 posts in this topic

Brian,

with the exception of Nimble, Starboard is the only hybrid vendor with a published ESRP benchmark that demonstrates the performance of the system.

While we appreciate Storage Review's work and help in finding some issues using your benchmark suites, I'd like to point out firstly, that the particular benchmark you're referring to is hardly a real-world benchmark (iometer running against a single 50GB file over SMB). Secondly, A similar test over iSCSI did not show any performance problems, so how could it be a caching tier deficiency? We have investigated this issue thoroughly and found it to be caused by the SMB daemon (samba) byte-level locking problem and have referred it to samba.org.

Also note that you were operating with a minimally configured system (an AC45), not the newer AC2000, 4000 or 4500 and ran most of your tests with the early release code.

Regards,

Kirill Malkin

Share this post


Link to post
Share on other sites

From a testing perspective, while we can agree that FIO as a synthetic benchmark isn't a perfect real-world workload, it does highlight areas of concern, which is why we still incorporate it into reviews with a growing number of application tests. As to not being a caching problem, if the platform advertises CIFS and iSCSI performance acceleration and only accelerates one of those and not the other, regardless of the core problem it still points to the caching process implemented.

EDIT: Just to clarify, the platform shipped to us is the AC72-1.

Share this post


Link to post
Share on other sites

Brian,

with the exception of Nimble, Starboard is the only hybrid vendor with a published ESRP benchmark that demonstrates the performance of the system.

While we appreciate Storage Review's work and help in finding some issues using your benchmark suites, I'd like to point out firstly, that the particular benchmark you're referring to is hardly a real-world benchmark (iometer running against a single 50GB file over SMB). Secondly, A similar test over iSCSI did not show any performance problems, so how could it be a caching tier deficiency? We have investigated this issue thoroughly and found it to be caused by the SMB daemon (samba) byte-level locking problem and have referred it to samba.org.

Also note that you were operating with a minimally configured system (an AC45), not the newer AC2000, 4000 or 4500 and ran most of your tests with the early release code.

Regards,

Kirill Malkin

Thanks for checking in, would have been great to have this conversation earlier, but I'm glad you stopped by today. We would have loved to get to the point of running VMmark and some of the other more relevant tests we have. We just never made it that far due to the issues highlighted. Starboard certainly helped to trailblaze the hybrid market, you guys deserve a lot of credit for the effort.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now
Sign in to follow this  
Followers 0