Posted by: Eric Siegel
As I initially discussed in my blog entries on 29 January of
this year ("SaaS
Network Diagnostics" and "SaaS
Network Performance SLAs"), the service level agreements (SLAs) and
diagnostic plans for Software as a Service (SaaS) are still in their infancy.
That blog entry discussed setting up measurement tools to provide SLA key
performance indicators and also to assist network operations detect and quickly
"triage" any performance problems, assigning the problem to the
correct organization for repair.
I've now decided, in light of the increasing interest in
SaaS and cloud computing (e.g., there's an entire track on
"Enterprise-ready Clouds" at Burton Group's Catalyst Conference
in Prague, April 19-22), I'll be writing a report on how to manage the
performance of cloud-based applications.
SO....
If anyone is willing to discuss the issue, please add some
comments here! How have YOU handled performance issues with your cloud / SaaS
provider? What is in the contract, and what happened when you had some
performance problems? How did the provider respond?
Hi Eric,
just to share one cent with you on this topic, I believe that one of the point to be solved by enterprise using SaaS is the management of the hybrid network (or hybrid traffic, according to deployment scheme. Actually, as branches will have to talk both to private DC (well localized and classically connected through the corporate VPN) AND to the 'virtual DC' somewhere behind the net, they find itself in the middle of a small hub, having to mitigate the different applications and networks in usage, which change significantly their classical 'spoke' situation. Here service and technology providers will have a lot to bring to guarantee that the most critical applications (which one, btw? SaaS or DC hosted?) will get their fair share of the resource. Cheers.
Thierry Grenot, CTO at Ipanema Technologies
Posted by: Thierry Grenot | March 08, 2010 at 11:20 AM