Abhijit Debnath’s Post

Data Migration and Replication to Oracle-->S3-->Snowflake using DMS: ------------------------------------------------------------------------------- If you’re considering migrating and replicating data from Oracle to Snowflake using DMS, it is essential to know the pros and cons of both the LogMiner and Binary Reader approaches. LogMiner: LogMiner is an SQL utility provided by Oracle that accesses both online and archived redo logs. It is commonly used for change data capture (CDC) in Oracle databases. While LogMiner offers several advantages, such as providing detailed information about changes (including SQL statements) and being useful for historical data analysis, it also has several disadvantages when compared to the Binary Reader approach. Binary Reader: Binary Reader is an AWS DMS feature that directly reads and parses the redo logs without using LogMiner. LogMiner Vs Binary Reader: 1. If you planning for high volume data changes then go for Binary Reader as it has lower impact on the Oracle source during migrations with a high volume of changes. 2. Binary Reader performs better with CDC for high volume changes compared to LogMiner. 3. Binary Reader supports CDC for LOB columns in Oracle while LogMiner does not support CDC if you are using Oracle 12C.  4. Binary Reader has less limitation on table char length and column char length. Currently, log miner has the limitation if the table char length or column char length is more than 30. How to Enable Binary Reader: It should be enabled at the endpoint level of DMS task. In extra connection attribute add below to enable it useLogminerReader=N;useBfile=Y In summary, choose LogMiner when you need detailed information and historical analysis, but consider Binary Reader for high volume migrations with minimal impact on the Oracle source.  If you are dealing with large volume migrations, Binary Reader might be the better choice.

If you’re in the process of selecting a distributed database, the following comparison matrix may be of assistance. It features three databases—YugabyteDB, Amazon Aurora, and CockroachDB—which have recently gained significant traction in various industries. The comparison I created below mainly aims to provide a clear distinction between YugabyteDB, Amazon Aurora, and CockroachDB. YugabyteDB Vs Aurora Vs CockroachDB Note: Below is the snapshot from my excel.

Have you tried Snowflake’s forecasting model? It’s user-friendly and straightforward. If you haven’t, let me provide a quick example. I created a basic time series table and applied the model to it. Here are the steps I followed to obtain a 99% accurate forecast for the next 10 days(Till 19th May). You can adjust the accuracy percentages as needed. Additionally, you can pass the query directly instead of using a view. drop table forecasting_test; create table forecasting_test(Product_count number, date_series timestamp_ntz(9)); insert into forecasting_test select UNIFORM(1, 10::NUMBER(10, 2), RANDOM()) PRODUCT_COUNT,  dateadd(   day,   '-' || row_number() over (order by null),   dateadd(day, '+1', current_date())  ) as date from table (generator(rowcount => 90)); select * from forecasting_test; GRANT USAGE ON DATABASE <DATABASE_NAME> TO ROLE ACCOUNTADMIN; GRANT USAGE ON SCHEMA <DATABASE_NAME>.<SCHEMA_NAME> TO ROLE ACCOUNTADMIN; GRANT CREATE SNOWFLAKE.ML.FORECAST ON SCHEMA <DATABASE_NAME>.<SCHEMA_NAME> TO ROLE ACCOUNTADMIN; ALTER ACCOUNT SET search_path='$current, $public, snowflake.ml'; show parameters like '%autocommit%' ; alter session set autocommit=TRUE; create or replace view vw_forecasting_test as select * from forecasting_test order by 2 desc; select * from vw_forecasting_test; describe view vw_forecasting_test; show snowflake.ml.forecast; create or replace snowflake.ml.forecast mdl_forecasting_test (   input_data => SYSTEM$REFERENCE('VIEW', 'vw_forecasting_test'),   timestamp_colname => 'DATE_SERIES',   target_colname => 'PRODUCT_COUNT' ); call <DATABASE_NAME>.<SCHEMA_NAME>.mdl_forecasting_test!forecast(   forecasting_periods => 10,   config_object => {'prediction_interval': 0.999} ); #HappyLearning

It takes immense excitement as I post this, to extend my heartfelt thanks to Snowflake for inviting me to the snowflake data for breakfast Toronto event, today. It was an absolute pleasure to be part of such an engaging and informative gathering. The combination of insightful discussions and delightful breakfast offerings made an incredible experience. I thoroughly enjoyed the opportunity to have a chat with industry leaders and fellow tech enthusiasts about the latest trends and developments. The event has left me feeling inspired and excited about the endless possibilities that Snowflake data presents. Thank you once again for organizing this fantastic event. I eagerly anticipate future opportunities to continue our exploration in this fascinating field. Snowflake Cognizant #SharedInvestigatorPlatform #DataForBreakfast #Toronto

I’m happy to share that I’ve obtained a new certification: SNOWPRO CORE CERTIFIED from Snowflake.

Well done. Proud of you

Kshouni Chakrabarti #Kudos I just wanted to say #ThankYou