We have SDE (10.3) running on SQL Server (SQL Server 2012 on Windows Server 2012 R2)
Over time we've noticed our GIS performance degrade, and can have good days and bad days. We run versioned editing, replication, have geometric and network datasets, and have archiving enabled. We run regular (several times a day) compresses to keep things running smoothly.
My question is whether having archiving enabled on our heavily edited datasets would have a noticible effect on performance? We are not in a position to test this, so wonder if anyone else has experience with testing the performance of archiving heavily edited data. Some of our feature classes could have anywhere from 100 to 10,000 edited records in a day.
We've had our geodatabase with all our different types of datasets and functions running for a number of years, and it's only since we migrated from arcsde 9.3 to 10.3 (about a year ago) that we've noticed the significant performance issues.