We are in process of upgrading Sitecore 7.2 to 8.2 Update 5 (may be Update 6 later). As part of the process client wants to migrate the historical analytics data from SQL to Mongo (On-Premise). Client have huge analytics database on 7.2 containing approx. 3+ years of data in it by having multi-site, multi-lingual etc.

Although we are planning to leverage the Batch Aggregation by configuring the MaximumBatchSize and MaxThread in History Worker process. But I strongly remember that in my previous Sitecore Upgrade, it took days to complete the data optimization, conversion and rebuilding the reporting database, even when they had less bulky analytics database than this.

Hence wanted to check with the community if we have any ways/guidelines to size and estimate (may be ballpark) for the Conversion process with various factors effecting.

  • Estimating the size of the xDB databases is very hard. I've done a few and we seem to see the MongoDBs are about 10-25% bigger than the SQL ones, but I think that depends on the nature of the data. I very strongly recommend running the migration tools against Azure DBs and/or VMs so you can scale up to SQL P16 DB and 64-CPU Mongos to speed up the job – Richard Hauer Dec 15 '17 at 8:00
  • Thanks Richard, but my question is related to estimate the efforts (Hours) a particular conversion may take considering the size of current Analytics database on 7.2. Cause we are on Agile and currently in sprint planning phase hence want to size the stories and tasks related to the conversion process. Hope it is making sense to you. – Amitabh Vyas Dec 15 '17 at 14:35
  • I would say “L” to “XL” – Richard Hauer Dec 15 '17 at 21:35