avi_vallarapu

Right. My point is - "How so many Services across Canada went down as they never designed their Services to overcome the network outage ? "

Anyways, I am more interested in how PostgreSQL can make us have trailing zeroes or no trailing zeroes depending on the type you choose or other work-arounds.

I think there is a way to get the inserted trailing zeroes in Oracle as well. So if you explicitly inserted 10.0500, you want to get 10.0500 but not 10.05 ?

Global variables inside an exception blog is one more interesting case when we see Oracle vs PostgreSQL

I worked on Oracle and MySQL for ages and seen tons of bugs. As a support company, the number of bugs we encounter with Postgres are none or hardly 1 in an year. I am still curios to see what those bugs are that made one scary !!!

Can you name a few such bugs please?

pg_amcheck can be helpful for validation if its a b-tree index

This is a tutorial on how to use the free Web tool to migrate Oracle Schema to PostgreSQL. Here is the URL for the tool.

http://ora2pg.migops.com/

pgBackRest is a wonderful backup and restore tool maintained by some of the Contributors of PostgreSQL. Kudos for making it so huge and feature rich.

This is actually a missing piece in Postgres. It was neatly blogged here. https://www.percona.com/blog/2020/05/21/failover-of-logical-...

PostgreSQL is the only database that won this title 4 times in the past 5 years.

Also, did you come across a situation where your Aurora bills were increasing with the amount of workload due to increase in IOPS ?

Did you experience a situation where you had to upgrade you Aurora Instances due to huge CPU utilization ?

No doubt in some aspects. Business logic may be great most of the times when it is out of the database. However, the theory cannot be applied to numerous legacy applications and also while migrating them all to PostgreSQL.

There may be several methods to upgrade Postgres, but this can be one of the most finest methods made available by Postgres community.

I have seen a lot of Customers still stuck on the Oracle versions where DBMS_JOB is heavily used.

Yes, good that Oracle added backward compatibility to DBMS_JOB thought there is DBMS_SCHEDULER in recent versions.

Yes, EnterpriseDB implemented it as part of their Commercial Postgres derived database : PPAS. It is not Open Source, whereas, PG_DBMS_JOB is Open Source and released under PostgreSQL license, which is the most liberal license :)

It is the extensibility of PostgreSQL that makes this possible.

Ya, especially when there exists PostgreSQL like Open Source databases that includes Native Partitioning, Replication, Full Backups plus Point in time recovery, Read Replicas, too many data types including boolean, uuid like types that are not possible with Oracle.

In that case, there even exist Orafce, a project that provides a lot of compatibility to a lot of Oracle Packages. For example, a lot of date functions, UTL_FILE, DBMS_OUTPUT and a lot more are already [art of this :)

https://github.com/orafce/orafce

As it is a full compatibility to Oracle's DBMS_JOB, it must help a lot while migrating from Oracle to PostgreSQL.

And this is mainly around array_dims being used and the arrays are not Associative Arrays indeed.

Definitely something everybody assumes and makes a mistake around.

So, it is just a Kudos to the PostgreSQL Community.

This is more of a sarcastic post that nothing can replace PostgreSQL :)

This backup tool even supports backups to Cloud with multiple repositories in parallel.

pgBackRest can also be considered as an Oracle RMAN equivalent and may be much more advanced features than Oracle RMAN.

While PLTCL is more lighter

PLTCL and PLPerl are 2 better approaches indeed.