MarkLogic Multi-Model Database: Enterprise Edition v. 11

My main use case for MarkLogic  is primarily for health and insurance companies, and secondarily for publishers who maintain a large set of books online and any company that deals with big data. We keep high volume of data in MarkLogic  because it provides a good search engine to query data in large datasets.

A specific example of how a publisher company would use MarkLogic in their workflow is maintaining data of a book, including who published that book, what the author name is, and how many editions exist. You can imagine that there are billions of books across the world, and we want to keep that digitized information in MarkLogic. If you want to search for a particular author, find how many books are published by a particular publisher, or discover editions of a particular book, you can get that information very quickly.

The best features MarkLogic offers include the ability to store unstructured data. You really don't need any particular schema, although it does support storing documents according to a particular schema, but that is not mandatory. You can store your data in XML and JSON format and you can store even binaries, JPEGs, PDFs and that kind of content in MarkLogic. The second feature is that it has a very rich search and cts APIs to build search engines on large datasets. The third is the spell suggestion APIs where, as you type, you will get suggestions of what you are searching for, which helps users understand what they are really searching for. Fourth is the redaction APIs. If you have a large set of data in a live environment and want redacted data in lower environments, the redaction APIs are very effective. From a user perspective, you can keep your data in redacted format in any environment, even on production.

The other valuable feature is the CPF, which is the Content Processing Framework . That is a very powerful tool. You can orchestrate how your document is processed before insertion or after insertion. A big example is that as soon as you have loaded a PDF into MarkLogic and want to extract the data from that PDF and store it as XML in MarkLogic for searching, that is where CPF is very useful.

MarkLogic has impacted my organization positively. We used to have one fetch API which was built using .NET API and the backend was SQL. We got to know from customers that the API was very slow and they were getting impacted. We moved to MarkLogic and created the API using JavaScript server-side language, and we saw almost 60% improvement in the speed of the search. The fetch results were 60% faster after that.

I think MarkLogic can be improved by providing good cloud infrastructure. Since we are in a very tech era, MarkLogic should provide good cloud infrastructure. If you look at other databases or systems like Kafka and MongoDB, they have cloud infrastructure. You do not need to worry about maintaining your own servers or provisioning your own servers. You simply log in and tell MarkLogic you want a certain number of clusters or nodes in a cluster and what cloud provider you want to use, then click okay, and they will build it for you. That is a headache with MarkLogic.

Apart from that, I don't see any issues. It is a really great product.

I have been using MarkLogic for 11 years.

MarkLogic is not entirely stable. We used to have production issues, mainly due to MarkLogic failures. The node goes down sometimes or some memory errors occur. It is stable, but if I rate it out of 100%, it is 80-85% stable. Every year we used to have four to five issues.

MarkLogic's scalability is very bad. In production, when you get to know that your data is increasing and you need to add one more node, that is not easy and not straightforward. You really need to do a lot of work and a lot of planning before you will be able to scale. But in today's world, if I talk about other systems like MarkLogic, if you have to scale, it is just a five-minute job. With MarkLogic, that is not the case.

I find customer support to be good.

MarkLogic has really good deployment processes. If you have a big MarkLogic architecture with hundreds of nodes in the cluster, many databases, forests, and many app servers, their deployment features like ML-Gradle can automate everything. You can create a new cluster within minutes through that Gradle. That is a very good feature they have provided.

I'm not much aware of pricing, setup costs, and licensing because that is another team that deals with it. But if I am correct, a few times I heard them saying that MarkLogic is quite costly, and they are looking to move away in the longer run for that reason.

Before choosing MarkLogic, we analyzed SQL and other solutions, but after analyzing and doing some POCs, we found that in our case, we had to store a large amount of binary data and SQL was not that good at storing binary data, especially when your need is to perform searches inside the binaries. That was something we did not find in SQL, and we did not prefer SQL for that reason.

My advice to others looking into using MarkLogic is that they really need to evaluate their use case. First, they need to evaluate what MarkLogic provides and what the features of MarkLogic are. Then they need to evaluate their use cases and check whether they are able to use the core features of MarkLogic. If not, then there is no point in moving to MarkLogic. Since it has a costly license, they really need to evaluate whether the features MarkLogic is providing would allow them to utilize most of them or not. I would love to see MarkLogic in more cloud-native architecture. I am giving this review a rating of 7.

My main use case for MarkLogic  is data development. We have a banking type of data, so we used to convert it to PDF or use some for applications we were running.

A quick, specific example of how I use MarkLogic  with my banking data involves creating PDFs based on XML files and developing subscription kind of applications. In that, we used to send scheduled PDFs and some scheduled emails generated from that application using MarkLogic.

In addition to my main use case, I also perform development and some MarkLogic admin work as well.

In my experience, the best features MarkLogic offers include indexing, which is a quick and efficient way to get the files and I really appreciate that.

The indexing feature of MarkLogic has helped my work by being quick, faster, and very helpful for the team to get the code quicker and to ensure everything is fast.

MarkLogic has positively impacted my organization by making everything quick and fast, and I believe that is a major change we have seen here. In fact, the development process is much quicker than other applications we were using previously.

As for areas where MarkLogic can be improved, nothing else comes to mind. I have been using MarkLogic for seven years, and many improvements have already been addressed. I do not recall any additional needed improvements.

I have been using MarkLogic for around seven years.

Previously, we were using SQL, but to save the PDFs and everything was not possible in SQL. Right now, it is possible in MarkLogic to save all the images, videos, and everything, which is very useful, and that is why we switched from SQL to MarkLogic.

I have seen a return on investment as time is really being saved using MarkLogic.

Regarding my experience with pricing, setup cost, and licensing, I find it quite high compared to other things, but I believe it is justified.

My advice for others looking into using MarkLogic is that it is a better way than SQL and while some main things can be improved, it is the best way right now in the market.

I chose a nine for my rating because it is faster, and I deducted one point because of the price of MarkLogic. My overall review rating is 9.