SoftGrid 4.5: The business case for Management Servers, Streaming Server or Stand-alone clients.
Has anyone attempted to calculate the ROI on implementing just a stand-alone client, the streaming server or a combination of both streaming servers & management servers or the heavy weight solution?
It looks to me the stand-alone client will be the winner in organizations with ESDs already in place.
Any thoughts?
It looks to me the stand-alone client will be the winner in organizations with ESDs already in place.
Any thoughts?
0 Comments
[ + ] Show comments
Answers (3)
Please log in to answer
Posted by:
kkaminsk
16 years ago
I think the heavyweight streaming server adds some cost because you need a SQL back end (heard MS upped the price), fault tolerant load balancing (you should be spending money here), and dedicated servers (can be VMs). If you have SMS you might be able to use your distribution points as streaming servers but I think in 4.5 SCCM will dual purpose distribution points with streaming servers. So in short yes, it costs some money to do streaming right. With that being said you can convert existing servers into lightweight streaming servers by installing IIS and the lightweight components. The only big issue is that you need the heavyweight streaming servers somewhere in the infrastructure for everything to function since they essentially hold the authentication mechanism.
I would like to point out that the SoftGrid server doesn't need to be dedicated hardware because.
a) The database doesn't need to be local.
b) The content folder doesn't need to be local.
Essentially the server becomes an OS + IIS but the load balancing is the big question. If you want to do it on the cheap use a VM with Microsoft Load Balancing in place and let it handle that function. Things can get a little expensive if you want to perform site to site SoftGrid fault tolerance with a Citrix Netscaler or equivalent Cisco product.
Now that I've ranted about options we need to talk ROI and that is the hard part. You can cut corners on the streaming infrastructure but where's the value!? ActiveUpgrade and ActiveDowngrade are huge features that allow you to seamlessly roll forward and roll backward application updates. This allows for infrastructure agility and can get you out of a tight spot if application testing was insufficient and the change needs to be backed out. As you can see it is very difficult to put a dollar figure on this feature. ActiveUpgrade allows for very minimal network impact when deploying a new application when combined with feature block one. Instead of delivering one gigabyte of data to a machine you can cut it down to 100-300 MB for feature block one (depends on app). This allows you to roll changes out in the middle of the day to your desktop environment which is a huge plus. Again, how do I put a dollar figure on this?
I think you have to figure out not from a dollar perspective what the ROI is but try to figure out if you want to leverage those features as part of your operations. Do you want that level of agility? Ok, then this is what that will cost. I guess it would be nice to show management some dollar figures because that is how they think but I don't think anyone has produced such numbers.
I would like to point out that the SoftGrid server doesn't need to be dedicated hardware because.
a) The database doesn't need to be local.
b) The content folder doesn't need to be local.
Essentially the server becomes an OS + IIS but the load balancing is the big question. If you want to do it on the cheap use a VM with Microsoft Load Balancing in place and let it handle that function. Things can get a little expensive if you want to perform site to site SoftGrid fault tolerance with a Citrix Netscaler or equivalent Cisco product.
Now that I've ranted about options we need to talk ROI and that is the hard part. You can cut corners on the streaming infrastructure but where's the value!? ActiveUpgrade and ActiveDowngrade are huge features that allow you to seamlessly roll forward and roll backward application updates. This allows for infrastructure agility and can get you out of a tight spot if application testing was insufficient and the change needs to be backed out. As you can see it is very difficult to put a dollar figure on this feature. ActiveUpgrade allows for very minimal network impact when deploying a new application when combined with feature block one. Instead of delivering one gigabyte of data to a machine you can cut it down to 100-300 MB for feature block one (depends on app). This allows you to roll changes out in the middle of the day to your desktop environment which is a huge plus. Again, how do I put a dollar figure on this?
I think you have to figure out not from a dollar perspective what the ROI is but try to figure out if you want to leverage those features as part of your operations. Do you want that level of agility? Ok, then this is what that will cost. I guess it would be nice to show management some dollar figures because that is how they think but I don't think anyone has produced such numbers.
Posted by:
SaiTech
16 years ago
Rating comments in this legacy AppDeploy message board thread won't reorder them,
so that the conversation will remain readable.
so that the conversation will remain readable.