Progress DB or Oracle DB ?

[rhi]

I am posting this in a new thread because I want to give everyone a chance to convince me why I should not choose an Oracle database over Progress database for my application, and I want to find out if there is a reason that I have not thought of, or am not aware of.

Assumptions:
- I already have an experienced Oracle DBA on staff that supports other applications already running on Oracle.
- I want to run Progress against an Oracle database because I want to take advantage of Oracle DB features. My application is 24/7.
- I have compensated for the different data types in Oracle and Progress and converted my code to run against an Oracle database.
- This can be either a new application requiring new licenses or an existing application.

What are the reasons that I should decide to migrate back to Progress backend Database?

 

[Lee Curzon]

I want to run Progress against an Oracle database because I want to take advantage of Oracle DB features.

Given that you seem to be talking about an (almost) all-Oracle environment, then surely you mean you want to run Progress against an Oracle database because you want to take advantage of Progress 4GL features?

Although to develop business logic would obviously be much simpler, the performance would be presumably be lousy (I've never used a Dataserver - so you'll have to confirm that), and there's no point using Progress as a Front End development system if you're not using a Progress database.

With the very restricted criteria you have supplied, I would recommend that you do NOT migrate back to Progress.

 

[rhi]

No, we want to take advantage of the Oracle DB. It is QAD's MFG/Pro applicaton, so 4gl is necessary. Dataserver performance is same or better based on tests already performed.

 

[Lee Curzon]

Sorry, I misunderstood. I thought you were baiting us again.

When you say performance is the same through the dataserver, do you mean that running MFG through a dataserver against Oracle gives the same performance as running MFG directly against Progress?

If so that means (off the top of my head) one or more of the following:

1. Oracle is faster than Progress.
2. MFG is badly written or outdated.
3. Your Progress system is badly configured.
4. a. Progress is bad at integrating with Oracle (or b. vice-versa).

Unfortunately, I cannot tell you which of those is the case (although I suspect you will claim 1 and 4a. But if most of your system is Oracle, I can understand why you wouldn't be keen on trying to integrate an old Progress application and database within it.

 

[rhi]

Sorry, I misunderstood. I thought you were baiting us again.

When you say performance is the same through the dataserver, do you mean that running MFG through a dataserver against Oracle gives the same performance as running MFG directly against Progress?

YES.

If so that means (off the top of my head) one or more of the following:

1. Oracle is faster than Progress.
2. MFG is badly written or outdated.
3. Your Progress system is badly configured.
4. a. Progress is bad at integrating with Oracle (or b. vice-versa).

Unfortunately, I cannot tell you which of those is the case (although I suspect you will claim 1 and 4a. But if most of your system is Oracle, I can understand why you wouldn't be keen on trying to integrate an old Progress application and database within it.

1. True, but reasons are due to DB features again. Example, when you start the DB you can designate which tables are pre-cached into memory. Eample 2, there are many database buffers for different operations, as opposed to only one (-B) in Progress. And many, many more.

2. True, but this can be a whole seperate discussion, and has nothing to do with databases

3. Not.

4a. Progress works very well w/ Oracle

4b. Not Applicable

 

[TomBascom]

I would think that your license cost would be a lot less if you weren't paying for a Progress dataserver plus an Oracle database.

That seems like a pretty good reason to me.

Other than that you seem to be an ideal candidate for using the Oracle dataserver. Go for it!

 

[rhi]

I'm not sure about that, but I am currently in the process of determing cost of the dataserver. However, last I knew, Progress new license cost was about $2500/user, and Oracle was $800/user. So if the dataserver costs something less than the difference ($1700), then it would be a break even. Added to this, my Oracle rep is willing to give a 44% discount, bringing the cost down to $448/user.

 

[TomBascom]

Progress cost per (database) user is nowhere near $2,500.

That might be the QAD license cost (which you're going to pay regardless) but the Progress database cost is nothing like that. (You may have been quoted a developers license? OpenEdge Studio or something?)

I don't have a current price sheet handy but the per user cost for Oracle Data Server lcenses used to be about 50% of an Enterprise DB user. And I think db seats are still in the 3 figure range. So adding that to your Oracle price gets yoiu pretty close to where the stand-alone Progress pricing ought to be . If the Progress rep gives you a similar discount then you should be way under the DataServer plus Oracle price.

 

[rhi]

Actually, this is my understanding, but it MAY be wrong. It's actually quite hard determinging this through QAD based on the way they license and charge, and I am actually in the middle the process of this cost breakdown and comarison.

Therefore, I should not have posted this cost comparison because it is probably not accurate. Sorry for that.

Costs aside for now, and other reasons for choosing Progress?

 

[TomBascom]

At some point everything comes back to cost...

I personally prefer fewer layers. The dataserver is, to me, a layer that adds no value so I'd avoid it on that basis. I also prefer to keep the number of parties and vendors to a minimum -- adding Oracle to the list of potential finger-pointers when something goes wrong doesn't entice me.

I can't directly comment on QAD's support but I'd want to check very carefully into whether ot not they support one choice better than the other. If 90% of their sales are on one configuration (maybe they are, maybe they aren't) then I'd expect problems just because you're different. Sort of like how 90% of your skills being Oracle centric makes you hostile to Progress

You probably aren't interested in or don't find credible assertions that Progress databases require less effort to take care of so I won't bother you with those reasons. Besides, that's a cost thing.


How about this one? If you're looking for an alternative to Oracle Progress is better than SQL Server :awink:

 

[BCM]

rhi -
Do you have other applications that are not written in Progress (for example: financial accounting, human resources, marketing, etc.)?
I have a preference to simple (fewer layers) as Tom stated. Wouldn't the business enterprise benefit from having compatible systems? What technology platform(s) are most closely aligned with your firm's business strategy? I believe it is better to maintain a high-level focus. Regardless of licensing costs/issues, the biggest cost and headache factors will result from trying to fit square pegs (no pun intended) in round holes. Try to stick with one approach that fits with your business strategy, organizational strategy (insource or outsource IT), and the successful technology choices already made at your firm.

 

[rhi]

To answer Tom & BCM,

The dataserver is no more of a "layer" for the Progress database than the Oracle database. The 4gl works much in the same way with either database, with some minor exceptions to the coding (ie. case sensitivity).

We have been using the dataserver to connect from MFG/Pro on Progress over to PeopleSoft on Oracle for about 6 years, and there has never been any problems.

We have 8 production systems in the enterprise, and MFG/Pro is the only one NOT on oracle, so to me the existence of the Progress database is the xtra layer in the stack at my company.

To Tom's point, cost is very relevant, and believe me, it is COSTLY to consistently have to take the Progress database down to perform various maintenance functions.

I used to defend the Progress DB like everyone else here seems to do, but that was back in the day when Progress was all I knew. Since I have learned what benefits Oracle can provide a company (and the DBA's who work for them), I have definitely switched teams.

Everyone here seems to only give me these "fringe" reasons (like xtra layers) to stick with Progress, and no real compelling reasons, like a database feature or performance stats, so I guess I have the answer.

 

[rhi]

You probably aren't interested in or don't find credible assertions that Progress databases require less effort to take care of so I won't bother you with those reasons. Besides, that's a cost thing.

By the way Tom, because the Progess database has to be taken down to do most maintenance, that forces me to work on weekends, when the cost of a downed production database is has less of an impact on the company.

How is this "less effort to take care of"???

Looking forward to your answer!!!:lol:

 

[BCM]

rhi -
Because all your applications run on Oracle except for Mfg/PRO which runs on Progress, I would, if possible, migrate the Mfg/PRO database to Oracle.

I know nothing of Mfg/PRO. Does the Mfg/PRO license include source code? If so, does your firm actively program in 4GL to customize Mfg/PRO?

If runtime 4GL code is executed against Oracle only and no 4GL development is performed then what minimal Progress licensing, if any, would be required at your firm? Would you be able to realize some cost savings in Progress licensing?

 

[joey.jeremiah]

i don't know if this thread is about practical suggestions?


depending on what it is you're trying to do, you could just upgrade

if you've got maintenance support, you can download it right now


while you're doing that maybe have a look at fathom

you could get a trial version, there's also a webcast on it

http://download.progress.com/open/products/pre-release/PSDN/fathom_overview_30a.zip

 

[rhi]

I trialed Fathom in house on my servers for 30 days.

Also trialed Oracle's Enterprise Manager Grid Control.

We bought the Oracle product, which I would recommend.

http://www.oracle.com/technology/products/oem/index.html

 

[joey.jeremiah]

is there something in particular you're trying to do ?

btw i know you're a dba, are you also a developer ?

 

[TomBascom]

You're obviously an Oracle heavy shop with a strong Oracle preference. You've also already got a Progress application that is Oracle ready.

I think you, personally, would be happier and better off with the Oracle dataserver. I'm certainly not going to stand in your way. In fact if you need any help I'd be happy to pitch in

But I wouldn't expect many other Progress customers to follow your lead. For the most part they don't share your circumstances or your opinions.