asked
No SQL, New SQL or Traditional RDBMS, what going to be the future??
Hello, As everybody aware of **Licensing cost and limitation of traditional RDBMS** (likes of Oracle and SQL Server). I am looking for an alternate for this costly databases as I have been asked to do so. I know about NOSQL and we can not use it mainly because of ACID requirement. Options which I see, either go with Postgres or go with NEW SQL (e.g. NUO DB) but since it is new, there aren't much/power user who can share their experience. Feature wise new sql seems promising but again their database engines are limited...therefore I am not sure in which DB I should go for?? Can any one guide me should I stick with SQL Server or should go with Postgres/ NEW SQL. Also I request, if any one have details or future prediction of NEW Sql, please share with me.
I'd be curious if you could list some of the specific limitations of "traditional RDBMS" that you are trying to work around. Based solely on your description, it sounds like the non "traditional" data stores are the ones with limitations that prevent their use in your environment. If you are considering any limitation other than licensing cost, PostgreSQL is solidly in the "traditional" camp and will be subject to the same limitations to which you alluded.
If you need ACID compliance, I'm with @KenJ here and I'm wondering why you'd be looking anywhere except at traditional structured storage. NewSQL is... new. New to the point of almost being vapor-ware. I sure wouldn't be putting my production systems on it just yet. If it's just licensing costs, PostgreSQL or MySQL seem to be pretty cheap and supply ACID compliance. I don't think you're going to get a good "answer" on this one, because it's really going to be down to opinion. I've never even heard discussion around most of the NewSQL databases you can see in a Google search.
@Kenj yes you are right about the same limitations in both DB but I was asked for why not Postgres and frankly speaking I don't have any solid answer to oppose them and specially when it comes on Cost. @GrantFritchey, I agree with you cause its really hard to find much detail and there is no such forum available for NewSQL, that's why I asked it here. Now about my requirements, DB should be ACID compliance, Low Cost and if needed scales out (since scaling up is always not possible in every condition). I don't have much experience in Postgres and NOSQL is just not the option, so its really hard for me to decide, in which direction should I go?
A cost that is frequently overlooked is knowledge. Yes, licensing is a cost, but so is a lack of knowledge. You save money licensing Postgres, but, if you haven't done lots of backups and restores and troubleshooting, if you hit a problem, and that's as likely as with any other system, you can't recover or can't recover in a timely manner. If you know SQL Server, that can make it much cheaper in terms of management costs. That's always something to keep in mind. Plus support, plus depth of hireable talent, plus, etc. Going with known quantities has a whole series of pluses. Unless you know you have a specific need for hard core scale out sharding style designs, I sure wouldn't make that one of the criteria.
17 People are following this question.
