PostgreSQL Plus Advanced Server says yes to NoSQL
February 02, 2016

PostgreSQL Plus Advanced Server says yes to NoSQL

Jamey Hanson | TrustRadius Reviewer
Score 10 out of 10
Vetted Review
Verified User
Review Source

Overall Satisfaction with Postgres Plus Advanced Server

We started recommending Postgres Plus Advanced Server (PPAS) to customers as a lower-cost alternative to Oracle Enterprise Edition, which it is. Now we also use PPAS for the simple and stable multi-master replication as well as the JSON capabilities native to PostgreSQL. We also have customers who use PostGIS, the geospatial add-on.
  • PPAS Oracle compatibility, especially the PL/SQL syntax, has made migrating database-tier code very simple. Most Oracle packages do not need to be changed at all and those that do are generally for simple reasons like a reserved word in PPAS that is allowed in Oracle.
  • PPAS xDB, the multi-master replication tool, is simple and - most important - does not break with network or other interruptions. We have been able to configure and forget, which our customers could never do with other multi-master tools.
  • Most people had no idea that PPAS and PostgreSQL have full CRUD support for JSON. They think you need a specialized product and/or that JSON is read-only. Every organization that I have worked with is evaluating adding JSON to their relational model.
  • There are too many custom add-ons that should be part of the default installation. For example, PostGIS - the geospatial tool, foreign-data-wrappers, XML & HSTORE data types, UUIDs and other pieces are not included by default.
  • The marketing and related material is too focussed on technology and not enough on capability. For example, you would not know that full CRUD on JSON is supported unless you read all the material - that is a big deal, but they seem to miss it.
  • The backup and recovery tools are not as good as Oracle's and they should be improved.
  • Simplified Oracle conversions because most of the PL/SQL does not need to be changed. This has allowed organizations that we support to do much of the conversion work themselves.
  • Allowed use of a single tool with mixed data types. For instance, no need to use a JSON-only product or a full-text-search tool because those are built in to PPAS.
  • Multiple customers added geo-spatial capabilities to their applications because PostGIS is simple and included.
  • Oracle Server
PPAS proved better for our customer's data-centric apps than Oracle in all but a few edge cases (encryption at rest and multi-TB database-tier backups) because it is simpler to install/maintain, runs nearly all Oracle-syntax SQL as well as ANSI SQL.
PPAS has much more JSON capabilities (full CRUD vs. read-only in Oracle), simpler geospatial, simpler / more stable replication and datatypes that match developer expectations, such as BOOLEAN and ENUMs.
Very well suited to applications mixing relational with NoSQL and/or complex data types, such as network analysis, JSON, key-value-pair, etc. Excellent if you have multiple programming languages in use (Python, Perl, R, etc.) because PPAS supports so many languages natively.

Not well suited if you need encryption at rest, compressed data or very large database-tier backups. [Large backups should be done at the storage tier if possible.]