Please share with the community what you think needs improvement with Oracle Database as a Service.
What are its weaknesses? What would you like to see changed in a future version?
The improvements we require are mostly regarding the documentation and compatibility with other systems or with other non-Oracle technologies. Documentation is difficult to find. Since we're experienced with Oracle technology, we have also found many errors in it. Not everything is 100% accurate. The solution should add functionalities, like database system functions, to speed up the query processing type. Any sort of functions or commands that would give you a faster result would be a welcome improvement. Microsoft Azure, for example, has some building functionalities to fetch big chunks of data and in a very fast way. I haven't seen a similar feature from Oracle.
I don't believe that any product is a perfect solution. I believe that each product has some minor disadvantage about which customers need to be able ready to work with or workaround. It may not be the fault of the product but lack of compatibility, the preferences of the user or lack of knowledge of existing features. I think these can all be the case in not maximizing what Oracle already has. Because of my preference for and knowledge of Toad and the way I work with Oracle, I am not up to date on all of the tools included in this product. But I think if we have all of the features needed as developers already in the product, Oracle will be much better. For example, something very important for all developers that could already be available now in Oracle products is weighting the data. This is a very important technical concern for data centers. The code that developers use is built up in some instructions, so I need to understand how each instruction is executed and how the current state of the code or data application is affecting the transactions. I found that this type of analysis was difficult in Oracle itself, so I used another product — like Toad — that would give me the ability to debug all the code instruction by instruction to really understand what is going on. It's quite an important feature. If it is available in Oracle now it is not very good or well described. This is a problem that needs a solution. If it is already there and the vendor does not know about it or understand it, I think there is a big issue. The difference is that to get the most out of the technology, the developers really need to know about the new features and how to find and use them. This is something about Oracle that I think is not very good. Along with that idea, I think that Oracle might do a little bit improve their interface as it is not very user-friendly and I have heard other developers say similar things. I didn't use myself because it would not help me as much or as quickly as other products to develop functions and to develop other features like reporting. I would have to learn a whole different way.
The direct scaling is a feature that has room for improvement. I would like to see virtual management machines that can scale storage well with zero downtime. The backup option and patching would then be much easier to do. I would also like to implement CDI with Database as a Service. There is also room for improvement with regards to scalability and availability. The maintenance is also an issue. When Oracle does maintenance, you need to be prepared because everything will be down.
When we troubleshoot, there are many tools that we need to use. One example is having to look at the system logs. The troubleshooting process should be improved so that we can resolve errors faster. Oracle demands a lot of memory and you need to have a strong machine. In the next release, I would like to see a framework for memory management.
Some of the technical features could be improved. They have a problem with some of the object types. I think this is one of the issues which is needed to be improved. If they could implement a managing tool that is better than the one they have now, that would be an improvement.
I would like to see a better dashboard for the storage. It need to be more complete overall. At the moment, it is basic.
Features that people want the most are already there. One of them is Oracle In-Memory which allows you to put things in a columnar data store in memory with Oracle's 12c (126.96.36.199+). You could have an in-memory database. You could have columnar data, which is compressed and in-memory (in the in-memory column store) or if you're doing transactional data, it's in the buffer cache (memory) so it's very fast to do that transactions. It's a different memory area altogether. When I look at the cloud, you'll need In-Memory more, because you're going to be dealing with larger volumes of data. The second feature that helps is multi-tenancy; Oracle already has this as of 12c Release 1. It is the ability to clone and move things around and maybe do an exchange partition (PDB - Pluggable Database) with one of your vendors or one of your customers out there (given security needs are met). Non-PDBs (or not using multi-tenancy is being deprecated). Leveraging and testing new features of 12.2 is also beneficial. Oracle 18c will be beneficial with Snapshot Carousel, PDB switchover, In-Memory External Tables and Polymorphic Tables. I think the way that Oracle and other vendors can move people quicker to the cloud is by educating people on the capabilities of the cloud and some of the benefits of the cloud. I think as customers and vendors out there and partners of Oracle look at providing solutions for the cloud, I think people will be more likely to go there, but what I think right now is the biggest hesitancy is there are a lot of other cloud providers don't have any of the features Oracle has. Oracle must let people know that if you have on-site Exadata, now you can have Exadata in the cloud. Easy to move. You've got a lot of data you want to archive? You can move it to a slower, larger server or a faster, smaller server depending on what you want to do. Just the capabilities that they have in-house or in the cloud. Some cost more than others, so effective planning is one big key to success. The cloud is gaining momentum, but the key to success is evaluating it fully and using appropriately.