January 20, 2016 at 1:33 pm
If you have someone good at it, finding the bad code and fixing it isn't as expensive as most would think and has the long term benefit of being rather bullet proof when it comes to performance and scalability.
Plus just throwing better faster hardware at the core problem just makes the statement that this is the best that we can do regarding speeding things up. This could give the message that the not so best practices in coding can continue because that is just something not given attention.
Plus you have to consider the costs of someone not wanting to run a report because it takes five minutes and it returns over 10, 000 rows just to get a few clients in there. I know it sounds extreme but I have seen these data dumps that have become the norm, there are intangible costs here. Oh well I am going off on a tangent here a little.
----------------------------------------------------
January 20, 2016 at 2:39 pm
Sometimes hardware just doesn't fix the issue though. We have a vendor app that is under performing we can't do anything to the DB or code as we upgraded from a 24core 64GB ram server running the DB off a 36spindle lun to a 64core 256GB server with an 80 disk lun and saw a 10-15% performance improvement which is shit for the capital involved.
Viewing 2 posts - 16 through 16 (of 16 total)
You must be logged in to reply to this topic. Login to reply