Posted: Tue Mar 25, 2008 8:38 pm Post subject: Is the problem solved?
I have a question concerning wether or not a solution should be considered permanent, or just as a workaround. Upon recieving the submitted problem, I failed to properly define the scope of it, and that has led to a "problem"(!). Please consider the following case:
Several incidents where logged as our users sometimes could not launch application "K". The application was virtualized and streamed to the users via SoftGrid.
A problem investigation was initiated, and it became clear quite quickly that the cause was that SoftGrid created way too large files for storing user settings, files which grew bigger by every launch.
Each time "K" was launched, SoftGrid would try to read the user settings before launching the application, and this could take anything from 2 minutes to several hours. The same thing happened when K was closed - SoftGrid would spend hours writing the file, and the application could not launch before the write-back had finished.
Only a few users had the problem in the beginning, but the impact increased steadily.
A workaround was identified and published: Simply delete the un-necessary user-settings file (testing showed no data was lost for the user), and a new (small) would be automatically created. It could take weeks before it reached a size when the problems began.
We put a great deal of effort and time into finding out WHY the files grew, but no conclusion was reached.
In the end, it was suggested and decided to discontinue the use of the SoftGrid viritualization/streaming software, and install the application locally on every server.
This means that a software update is a lot more work for our technical staff, but the users do no longer have this issue, and the incidents have stopped.
On to the final question:
Our technical staff consider this a workaound. In their opinion, SoftGrid should be re-deployed as soon as the cause of the growing files is located and eliminated in a test environment, and they want Problem Management to handle this ("You didnt' solve the REAL problem. Now we have more work than before").
In my opinion, the problem is solved and should be closed: A cause was identified, an RFC was created and the change is implemented and verified. The users are no longer experiencing disruptions to the service, and an eventual re-deployment of SoftGrid should go back to the design fase. It is futile spending even more resources on this than we already have, with no ending in sight.
How should this case be treated?
All opinions and explenations would be deeply appreciated.
The Norwegian correctional services IT-centre
Joined: Mar 04, 2008 Posts: 1892 Location: Helensburgh
Posted: Tue Mar 25, 2008 9:48 pm Post subject:
I think the best way to tackle this situation is to take a step back and get the wider perspective into focus. Don't debate whether it's a workaround or a resolution at all. Examine the pros and cons of the two delivery solutions in their entirety.
Firstly, does policy or strategic direction have a bearing? If so, you have to find a solution that conforms or you have to review your policy or strategy. (Perhaps the available tools are not sufficiently mature for what you want)
Secondly, what about costs and benefits of the two approaches. Include here licenses, resources, availability of skills for support and maintenance, reliability, future directions ...
Thirdly, what are the costs and risks and probabilities in trying to re-implement your original solution? How sure are you that the cause of the defect can be identified and corrected and how long will it take and what will be the cost?
There need to be discussions to clarify these questions and come to a decision on the way forward. Then you can sensibly decide the most appropriate management approach - problem management, project management or whatever, according to your available resources and ways of working. _________________ "Method goes far to prevent trouble in business: for it makes the task easy, hinders confusion, saves abundance of time, and instructs those that have business depending, both what to do and what to hope."
William Penn 1644-1718
Joined: Oct 07, 2007 Posts: 441 Location: Jakarta, INA
Posted: Tue Mar 25, 2008 11:51 pm Post subject:
I think I can understand the technical people's concern.
By using virtualization, the infrastructure can be made efficient since there will be no dedicated server for specific application.
The resolution or workaround of installing "K" locally would create a complex infrastructure, and in a long run it would affect the infrastructure's performance. Then there might be inbalance problem in capacity planning, and so on and so on.
I think they desperately need the virtualization, maybe you can recommend using another virtualization tool?
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum