Working on a plan to Migrate our Apps to Servoy if current testing and development works out.
It would make sense, perhaps, at some point to swap out our production SQL server to Postgres. Currently there are only a few modules where SP’s are used and we do not use views or triggers so migration should be relatively easy.
There are tables with millions of records involved.
has anyone had bad experiences with Postgres in regards to Stability, Maintainability, Recoverability etc etc.
johnh:
has anyone had bad experiences with Postgres in regards to Stability, Maintainability, Recoverability etc etc.
Short answer:
I’ve been using PostgreSQL since 7.4 and I never had any bad experiences.
Long answer:
PostgreSQL always has been rock solid for me and actually always had the name to be rock solid and it got MUCH faster from 8.3 and up. In fact with 9.1 it became very feature complete and even got some innovations that are first- or were in the race to be the first in the industry, let alone in the OpenSource world.
PostgreSQL can definitely go head to head with any commercial offering that is out there and yes I am talking about Oracle, Microsoft and IBM (the big 3).
And if there are any bugs that are critical the community puts out a bugfix version in days.
The fact that it’s OpenSource is actually a big selling point. Not because it’s FREE but because of the incredible involved and technically skilled community (and companies like NTT, EnterpriseDB, VMWare, Fujitsu, etc.) behind it. One that is hard to match by any commercial vendor, expecially for the price .
Yes, I make daily backups of my production database myself. But you could also setup a hot standby server for failover using (the built-in) streaming replication.
There are many high-availability scenarios you could implement.
But in any scenario you should always make an offsite backup be it streaming or by dumps, which holds true for any vendor.
In development over several years I have used Sybase SQLAnywhere, MS SQL, MySQL and PostgreSQL, migrating data between them. Servoy works with them all. Make sure you check out every procedure to ensure your original code does not contain db specific queries which are not recognised properly in the target db.
In brief: I support Robert’s experience.
In production I have found PostgreSQL to be rock solid and fast for the size of database in use: which in my client’s case is around a million records. Available maintenance and deployment tools are sufficiently good for both development and production. The db makes backups of each of two databases automatically every weekday evening to a backup folder on the production server. A scheduled task in the early morning copies the backups over the LAN to a separate dedicated PC. Offsite backups are made once a week - which is fine for my client’s risk profile.