No impact other than that you will no longer be able to take log backup and hence will not be able to restore to a point-in-time after switching to simple. Also if you are running database mirroring you will not be able to switch from full recovery
You can shrink the log to 50MB, but if the DB is active there's every chance it will grow again. 50MB is very small for a transaction log for an active database.
How much of a performance hit (a little, moderate, A LOT, don't do it, etc.) is there from shrinking the log file? The reason I ask is I'm faced with a situation where a production database has a much larger log file than is necessary. The database in question is a production database and it is pretty much used 24/7, so I'm attempting to find out if it is 'okay' or feasible to perform the action of shrinking a 100 GB .ldf during production hours.
The following was suggested to me in order to fix the matter:
1. Backup the TLOG
2. Change DB from Full to Simple
3. Change the initial size of the TLOG to something much smaller
4. Perform the log shrink
5. Change from Simple to Full
6. Perform full backup of DB to start log chain
Is this proper? If not, what would your process be to fix this matter?