You might need a Technical Audit and Performance Evaluation. There shouldn't be a significant decrease in performance to the degree that you're stating - BUT :
1. If you're using EXACTLY the same hardware between the "baremetal" instance compared to a Hyper-V instance, you will definitely see a performance tradeoff (approximately 20% based on benchmarks I published about 6 years ago).
2. You're not comparing apples to apples - it sounds like you went from Windows 2008 to Windows 2012, and upgraded the Tools Release from 8.98 to 9.1.5.x - so you've made some changes to E1. Check logs, and ensure that everything is as "fast" as possible !
3. When you refer to the UBE's slowing down - are these long running UBE's ? ie, are you having an issue with, say, a UBE running 1 hour in your baremetal and 2 hours in the virtual environment ? Or are you looking at UBE's that took 30 seconds and now take 1 minute ? Short running UBE's taking longer might be indicative of a JDE.INI change - but long running UBE's are indicative of overall performance issues. Need some clarity on this.
4. What did you do with your Database ? Is it running Oracle or SQL Server ? What versions ? Where are the disks located ? Most performance issues end up being a database performance issue - for example, the database files are placed onto a RAID 5 disk instead of RAID 10 or something similar. If you're placing the files onto virtual disks - then that could be an issue as well. You might be seeing good performance with processes that can be retained in memory - but as soon as disk access occurs, you might have latency issues.
So, the answer is that you need to investigate more thoroughly, provide more details - and, if you're seriously concerned about performance in a production instance, hire a specialist !
