Pentaho Data Integration and Analytics Reviews

We're using it for data warehousing. Typically, we collect data from numerous source systems, structure it, and then make it available to drive business intelligence, dashboard reporting, and things like that. That's the main use of it. 

We also do a little bit of moving of data from one system to another, but the data doesn't go into the warehouse. For instance, we sync the data from one of our line of business systems into our support help desk system so that it has extra information there. So, we do a few point-to-point transfers, but mainly, it is for centralizing data for data warehousing.

We use it just as a data integration tool, and we haven't found any problems. When we have big data processing, we use Amazon Redshift. We use Pentaho to load the data into Redshift and then use that for big data processing. We use Tableau for our reporting platform. We've got quite a number of users who are experienced in it, so it is our chosen reporting platform. So, we use Pentaho for the data collection and data modeling aspect of things, such as developing facts and dimensions, but we then publicly export that data to Redshift as a database platform, and then we use Tableau as our reporting platform.

I am using version 8.3, which was the latest long-term support version when I looked at it the last time. Because this is something we use in production, and it is quite core to our operations, we've been advised that we just stick with the long-term support versions of the product.

It is in the cloud on AWS. It is running on an EC2 instance in AWS Cloud.

It enables us to create low-code pipelines without custom coding efforts. A lot of transformations are quite straightforward because there are a lot of built-in connectors, which is really good. It has got connectors to Salesforce, which makes it very easy for us to wire up a connection to Salesforce and scrape all of that data into another table. Their flows have got absolutely no code in them. It has a Python integrator, and if you want to go into a coding environment, you've got your choice of writing in Java or Python.

The creation of low-code pipelines is quite important. We have around 200 external data sets that we query and pull the data from on a daily basis. The low-code environment makes it easier for our support function to maintain it because they can open up a transformation and very easily see what that transformation is doing, rather than having to troll through reams and reams of code. ETLs written purely in code become very difficult to trace very quickly. You spend a lot of time trying to unpick it. They never get commented on as well as you'd expect, whereas, with a low-code environment, you have your transformation there, and it almost self documents itself. So, it is much easier for somebody who didn't write the original transformation to pick that up later on.

We reuse various components. For instance, we might develop a transformation that does a lookup based on the domain name to match to a consumer record, and then we can repeat that bit of code in multiple transformations. 

We have a metadata-driven framework. Most of what we do is metadata-driven, which is quite important because that allows us to describe all of our data flows. For example, Table one moves to Table two, Table two moves to table three, etc. Because we've got metadata that explains all of those steps, it helps people investigate where the data comes from and allows us to publish reports that show, "You've got this end metric here, and this is where the data that drives that metric came from." The variable substitution that Pentaho has to allow metadata-driven frameworks is definitely a key feature that Pentaho offers.

The ability to automate data pipeline templates affects our productivity and costs. We run a lot of processes, and if it wasn't reliable, it would take a lot more effort. We would need a lot bigger team to support the 200 integrations that we run every day. Because it is a low-code environment, we don't have to have support instances escalated to the third line support to be investigated, which affects the cost. Very often our support analysts or more junior members are able to look into what an issue is and fix it themselves without having to escalate it to a more senior developer.

The automation of data pipeline templates affects our ability to scale the onboarding of data because after we've done a few different approaches and we get new requirements, they fit into a standard approach. It gives us the ability to scale with code and reuse, which also ties in with the metadata aspect of things. A lot of our intermediate stages of processing data are purely configured in metadata, so in order to implement transformation, no custom coding is required. It is really just writing a few lines of metadata to drive the process, and that gives us quite a big efficiency.

It has certainly reduced our ETL development time. I've worked at other places that had a similar-sized team to manage a system with a much lesser number of integrations. We've certainly managed to scale Pentaho not just for the number of things we do but also for the type of things we do.

We do the obvious direct database connections, but there is a whole raft of different types of integrations that we've developed over time. We have REST APIs, and we download data from Excel files that are hosted in SharePoint. We collect data from S3 buckets in Amazon, and we collect data from Google Analytics and other Google services. We've not come across anything that we've not been able to do with Pentaho. It has proved to be a very flexible way of getting data from anywhere.

Our time savings are probably quite significant. By using some of the components that we've already got written, our developers are able to, for instance, put in a transformation from a staging area to its model data area. They are probably able to put something in place in an hour or a couple of hours. If they were starting from a blank piece of paper, that would be several days worth of work.

The graphical nature of the development interface is most useful because we've got people with quite mixed skills in the team. We've got some very junior, apprentice-level people, and we've got support analysts who don't have an IT background. It allows us to have quite complicated data flows and embed logic in them. Rather than having to troll through lines and lines of code and try and work out what it's doing, you get a visual representation, which makes it quite easy for people with mixed skills to support and maintain the product. That's one side of it. 

The other side is that it is quite a modular program. I've worked with other ETL tools, and it is quite difficult to get component reuse by using them. With tools like SSIS, you can develop your packages for moving data from one place to another, but it is really difficult to reuse a lot of it, so you have to implement the same code again. Pentaho seems quite adaptable to have reusable components or sections of code that you can use in different transformations, and that has helped us quite a lot.

One of the things that Pentaho does is that it has the virtual web services ability to expose a transformation as if it was a database connection; for instance, when you have a REST API that you want to be read by something like Tableau that needs a JDBC connection. Pentaho was really helpful in getting that driver enabled for us to do some proof of concept work on that approach.

Although it is a low-code solution with a graphical interface, often the error messages that you get are of the type that a developer would be happy with. You get a big stack of red text and Java errors displayed on the screen, and less technical people can get intimidated by that. It can be a bit intimidating to get a wall of red error messages displayed. Other graphical tools that are focused at the power user level provide a much more user-friendly experience in dealing with your exceptions and guiding the user into where they've made the mistake.

Sometimes, there are so many options in some of the components. Some guidance about when to use certain options embedded into the interface would be good so that people know that if they set something, what would it do, and when should they use an option. It is quite light on that aspect.

I have been using this solution since the beginning of 2016. It has been about seven years.

We haven't had any problems in particular that I can think of. It is quite a workhorse. It just sits there running reliably. It has got a lot to do every day. We have occasional issues of memory if some transformations haven't been written in the best way possible, and we obviously get our own bugs that we introduce into transformations, but generally, we don't have any problems with the product.

It meets our purposes. It does have horizontal scaling capability, but it is not something that we needed to use. We have lots of small-sized and medium-sized data sets. We don't have to deal with super large data sets. Where we do have some requirements for that, it works quite well. We can push some of that processing down onto our cloud provider. We've dealt with some of such issues by using S3, Athena, and Redshift. You can almost offload some of the big data processing to those platforms.

I've contacted them a few times. In terms of Lumada's ability to quickly and effectively solve issues that we brought up, we get a very good response rate. They provide very prompt responses and are quite engaging. You don't have to wait long, and you can get into a dialogue with the support team with back and forth emails in just an hour or so. You don't have to wait a week for each response cycle, which is something I've seen with some of the other support functions. 

I would rate them an eight out of 10. We've got quite a complicated framework, so it is not possible for us to send the whole thing over for them to look into it, but they certainly give help in terms of tweaks to server settings and some memory configurations to try and get things going. We run a codebase that is quite big and quite complicated, so sometimes, it might be difficult to do something that you can send over to show what the errors are. They wouldn't log in and look at your actual environment. It has to be based on the log files. So, it is a bit abstract. If you have something that's occurring just on a very specific transformation that you've got, it might be difficult for them to drill into to see why it is causing a problem on our system.

I have a little bit of experience with AWS Glue. Its advantage is that it is tied natively into the AWS PySpark processing. Its disadvantage is that it writes some really difficult-to-maintain lines of code for all of its transformations, which might work fine if you have just a dozen or so transformations, but if you have a lot of transformations going on, it can be quite difficult to maintain.

We've also got quite a lot of experience working with SSIS. I much prefer Pentaho to SSIS. The SSIS ties you rigidly to your data flow structure that exists at design time, whereas Pentaho is very flexible. If, for instance, you wanted to move 15 columns to another table, in SSIS, you'd have to configure that with your 15 columns. If a 16th column appears, it would break that flow. With Pentaho, without amending your ETL, you can just amend your end data set to accept the 16th column, and it would just allow it to flow through. This and the fact that the transformation isn't tied down at the design time make it much more flexible than SSIS.

In terms of component reuse, other ETL tools are not nearly as good at being able to just pick up a transformation or a sub-transformation and drop it into your pipelines. You do tend to keep rewriting things again and again to get the same functionality.

I was here during the initial setup, but I wasn't involved in it. We used an external company. They do our upgrades, etc. The reason for that is that we tend to stick with just the long-term support versions of the product. Apart from service packs, we don't do upgrades very often. We never get a deep experience of that, so it is more efficient for us to bring in this external company that we work with to do that.

It is always difficult to quantify a return on investment for data warehousing and business intelligence projects. It is a cost center rather than a profit center, but if you take the starting point as this is something that needs to be done, you could pick up the tools to do it. In the long run, you would necessarily find that they are much cheaper. If you went for more of a coded approach, it might be cheaper in terms of licensing, but then you might have higher costs of maintaining that.

It does seem a bit expensive compared to the serverless product offering. Tools, such as Server Integration Services, are "almost" free with a database engine. It is comparable to products like Alteryx, which is also very expensive.

It would be great if we could use our enterprise license and distribute that to analysts and people around the business to use in place of Tableau Prep, etc, but its UI is probably a bit too confusing for that level of user. So, it doesn't allow us to get the tool as widely distributed across the organization to non-technical users as much as we would like.

I would advise taking advantage of using metadata to drive your transformations. You should take advantage of the very nice and easy way in which variable substitution works in a lot of components. If you use a metadata-driven framework in Pentaho, it will allow you to self-document your process flows. At some point, it always becomes a critical aspect of a project. Often, it doesn't crop up until a year or so later, but somebody always comes asking for proof or documentation of exactly what is happening in terms of how something is getting to here and how something is driving a metric. So, if you start off from the beginning by using a metadata framework that self documents that, you'll be 90% of the way in answering those questions when you need to.

We are satisfied with our decision to purchase Hitachi's products, services, or solutions. In the low-code space, they're probably reasonably priced. With the serverless architectures out there, there is some competition, and you can do things differently using serverless architecture, which would have an overall lower cost of running. However, the fact that we have so many transformations that we run, and those transformations can be maintained by a team of people who aren't Python developers or Java developers, and our apprentices can use this tool quite easily, is an advantage of it.

I'm not too familiar with the overall roadmap for Hitachi Vantara. We're just using the Pentaho data integration products. We don't use the metadata injection aspects of Pentaho mainly because we did have a need for them, but we know they're there. 

I would rate it a seven out of 10. Its UI is a bit techy and more confusing than some of the other graphical ETL tools, and that's where improvements could be made.

We used it for ETL to transform data from flat files, CSV files, and database. We used PostgreSQL for the connections, and then we would either import it into our database if the data was in from clients, or we would export it to files if clients wanted files or if a vendor needed to import the files into their database.

The biggest benefit is that it's a low-code solution. When you hire junior ETL developers or engineers, who may have a schooling background but no real experience with ETL or coding for ETL, it's a UI-based, low-code solution in which they can make something happen within weeks instead of, potentially, months.

Because it's low-code, while I could technically have done everything in Python alone, that would definitely have taken longer than using Pentaho. In addition, by being able to standardize pipelines to handle the onboarding process for new clients, development costs were significantly reduced. To put in perspective, prior to my leading the effort to standardize things, it would typically take about a week to build a feed from start to finish, and sometimes more depending on how complicated it was. With this solution, instead of it taking a week, it was reduced to an afternoon, or about three hours. That was a significant difference.

Instead of paying a developer a full week's worth of work, which could be $2,500 or more, it cut it down to three hours or about $300. That's a big difference.

The fact that it's a low-code solution is valuable. It's good for more junior people who may not be as experienced with programming. In our case, we didn't have a huge data set. We had small and medium-sized data sets, so it worked fine.

The fact that it's open source is also helpful in that, if a junior engineer knows they are going to use it in a job, they can download it themselves, locally, for free, and use test data to learn it.

My role was to use it to write one feed that could facilitate multiple clients. Given that it was an open-source, free solution, it was pretty robust in what it could do. I could make lookup tables and databases and map different clients, and I could use the same feed for 30 clients or 50 clients. It got the job done for our use case.

In addition, you can install it wherever you need it. We had installed versions in the cloud and I also had local versions.

If you're working with a larger data set, I'm not so sure it would be the best solution. The larger things got the slower it was.

It was kind of buggy sometimes. And when we ran the flow, it didn't go from a perceived start to end, node by node. Everything kicked off at once. That meant there were times when it would get ahead of itself and a job would fail. That was not because the job was wrong, but because Pentaho decided to go at everything at once, and something would process before it was supposed to. There were nodes you could add to make sure that, before this node kicks off, all these others have processed, but it was a bit tedious. 

There were also caching issues, and we had to write code to clear the cache every time we opened the program, because the cache would fill up and it wouldn't run. I don't know how hard that would be for them to fix, or if it was fixed in version 10.

Also, the UI is a bit outdated, but I'm more of a fan of function over how something looks.

One other thing that would have helped with Pentaho was documentation and support on the internet: how to do things, how to set up. I think there are some sites on how to install it, and Pentaho does have a help repository, but it wasn't always the most useful.

I used Hitachi Lumada Data Integration (Pentaho) for three years

In terms of the stability of the solution, as I noted, I wouldn't use it for large data sets. But for small to midsize companies that are looking for a low-code solution that isn't going to break the budget, it's a great tool for them to use.

It worked and it was stable enough, once we figured out the little quirks and how to get around them. It mostly handled our production workflows without issue.

I think it could scale, but only up to a point. I didn't test it on larger datasets. But after talking to people who have worked on larger datasets, they wouldn't recommend using it, but that is hearsay.

In my former company, there were about five people in the data engineering department who were using the solution in their roles as ETL data integration Specialists.

In that company, it's their go-to solution and I think it will work for everything that they need. When I was there, I tried opening pathways to different things, but there were so many feeds already on it, and it worked for what they need, and it's low-code and open source, so I think they'll stick with it. As they gain more clients they'll increase their usage of it.

The initial setup wasn't that complicated. You have to set the job environment variables and that was probably the most complicated part, and would be especially so if you're not familiar with it. Otherwise, it was just a matter of downloading the version needed, installing it, and learning how to use the different components. Overall, it was pretty easy and straightforward.

The first time we deployed it, not knowing what we were doing, it took a couple of days, but that was mainly troubleshooting and figuring out what we were doing wrong because we hadn't used it before. After that, it would take maybe 30 minutes or an hour.

In terms of maintenance for Pentaho, one developer per feed is what is typically assigned. It will depend on the workflow of the company and how many feeds are needed. In our case there were five people involved.

It saved us a lot of money. Given that it's open source, and the amount of time over the three that I used it, and the fact that they were using it several years prior, means a lot of money was definitely saved by using Pentaho versus something else.

If a company is looking for an ETL solution and wants to integrate it with their tech stack but doesn't want to spend a bunch of money, Pentaho is a good solution. SSIS cores were $10,000 a piece. Although I don't know what they cost nowadays, they're expensive. 

Pentaho is a nice option without having to pay an arm and a leg. We even had a complicated data set and Pentaho was able to handle pretty much every type of scenario, if we thought about it creatively enough. I would recommend it for a company in that position.

While the capabilities of Pentaho are good enough for light work, I've started using Alteryx Designer, and it is so much more robust in everything that you can do in real time. I've also used SSIS.

When you run something in Pentaho, you can click on it to see the output of each one, but it's hard to really change anything. For example, if I were to query data from a database and put it into a "select," if I wanted to reorganize within the select based on something like the first initial of someone's name, it provided that option. But when I would do it, sometimes it would throw an error and I'd have to run the feed again to see it.

The nodes, or the components, in Pentaho can probably do about 70 percent of what you can do in Alteryx. Don't get me wrong, Pentaho worked for what we needed it for, with just a few quirks. But as a data engineer, I'm always interested in and excited to work with new technologies that may offer different benefits. In this case, one of the benefits is that each node in Alteryx has many more capabilities in real time. I can look at the data that's coming into the node and the data that's going out. There was a way to do that in Pentaho, if you right-clicked and looked, but it would tell you the fields that were coming in and out and not necessarily the data. It's nice to be able to troubleshoot, on the spot, node-by-node, if you're having an issue. You can do that easily with Alteryx.

In addition to being able to look at data coming in and out of the node, you can also sort it easily and filter it within each data node in Alteryx, and that is something you can't do in Pentaho.

Another cool thing with Alteryx, although it's a very small difference, is that you don't have to save the workflow before you run it. Pentaho forces you to do that. Of course, it's always good to save.

A good thing about Pentaho is that it's not that hard to learn, from an ETL perspective. The way that Pentaho has things laid out they are pretty intuitively organized in the panel: Your input—flat file, CSV, or database—and then the transformation nodes. 

It was a good baseline and a good open-source tool to use to learn ETL. It's good to have exposure to multiple tools because every company has different needs and, depending on their needs, it would be a different recommendation.

The lessons I learned using it: Make sure you clear the cache when you open the program. Also, if there are any critical points in your flow that are dependent upon previous nodes, make sure that you put blocking steps in. Make sure you also set up the job environment variables correctly, so that Pentaho runs.

It worked for what we did but, personally, I wouldn't use it. In the new company I'm working for, we are using large financial data sets and I'm not so sure it could handle that. I know there's an Enterprise version, but I didn't use that.

The solution can handle ingestion through to export, but you still have to have a batch or Python script to run it with an automation process. I don't know if the Lumada version has something different, but with what I was using, you were simply building the pipeline, but the pipeline outside of the program had to be scheduled and run, and we had other tools to check that the output was as expected.

We used version 7 for a while and we were reluctant to upgrade to version 9 because we had an 834 configuration, meaning a government standardized feed that our developer spent two years building. There was an issue whenever we tried to run those feeds on version 9, so we were reluctant to upgrade because things were working on 7. We ended up finding out that it didn't take much work for us to fix the problem that we were having with version 9 and, eventually, we moved to it. With every version upgrade of anything, there are going to be pros and cons.

Depending on what someone needs it for, if it's a small project and they don't want to pay for an enterprise solution, I would recommend it and give it a nine out of 10. The finicky things were a little frustrating, but the fact that it's free, can be deployed easily, and that it can fulfill a lot of things on a small scale, are plusses. If it were for a larger company that needed an enterprise solution, I wouldn't recommend it. In that case, it would be one out of 10.

For a smaller company or one with a smaller budget, a company that doesn't have highly complex ETL needs, Pentaho is definitely a great option. If a company has the budget and has really specific needs and large data sets, I would suggest looking elsewhere.

The use case is for data ETL on our various data repositories. We use it to aggregate and transform data for visualization purposes for our upper management.

Currently, I am using the PDI locally on my laptop, but we are undergoing an integration to push this off. We have purchased the Enterprise edition and have licenses, and we are just working with our infrastructure to get that set up on a server. 

We haven't yet launched the Enterprise edition, so I've had very minimal touch with Lumada, but I did have an overview with one of the engineers as to how to use the customer portal in terms of learning documentation. So, the documentation and support are basically the two main areas that I've been using it for. I haven't piped any data or anything through it. I've logged in a couple of times to the customer portal, and I've pretty much been using it as support functionality. I have been submitting requests to understand more about how to get everything to be working for the Enterprise edition. So, I have been using the Lumada customer portal mostly for Pentaho Data Integration.

When we get a question from our CEO that needs a response and that requires a little bit of legwork of pulling in from various market data, our own in-house repositories, and everything else, it allows me to arrive at the solutions much faster than having to do it through scripting in Python, coding, or anything else. I use multiple tools within my toolkit. I'm pretty heavy on Python, but I find that I can do quite a bit of pre-transformation of the data within the actual application for PDI Spoon than having to do everything through coding in Python.

It has significantly reduced our ETL development time. I can't really quantify the hours, but it's a no-brainer for me for just pumping in things. If I have a simple question to ascertain, I can pull up and create any type of job or transform to easily get the solution within minutes, as opposed to however many hours of coding it would take. My estimate is that per week, I would be spending about 75% of my time in coding external to the application, whereas, with the application itself, I can do things within a fraction of that. So, it has reduced my time from 75% to about 5%. In terms of the cost of full-time employee coding and everything, the savings would also roughly be the same, which is from 75% to 5% per week. There is also a broader impact on other colleagues within my team. Currently, their processes are fairly manual, such as Excel-based, so the time savings are carried over to them as well.

I'm at the early stages with Lumada, and I have been using the documentation quite a bit. The support has definitely been critical right now in terms of trying to find out more about the architectural elements that need to go in for pushing the Enterprise edition.

I absolutely love Hitachi. I'm one of the forefront supporters of Hitachi for my firm. It's so easy to integrate within our environments. In terms of being able to quickly build ETL jobs, transform, and then automate them, it's really easy to integrate throughout for data analytics. 

I also appreciate the fact that it's not one of the low-code/no-code solutions. You can put as much JavaScript or another code into it as you want, and that makes it a really powerful tool.

I haven't been able to broach all the functionality of the Enterprise edition because it hasn't been integrated into our server. We're still building out the server, app server, and repository to support it.

In the Community edition, it would be nice to have more modules that allow you to code directly within the application. It could have R or Python completely integrated into it, but this could also be because I'm using an older version.

I have been using it here for about two months. 

I haven't had any problems with stability. Right now, for the implementation of the Enterprise edition, we're trying to make sure that it's highly available in case anything goes down, and we have proper safety nets in place, but personally, I haven't found any issues.

It seems highly scalable. I've used the product in other firms, and we've managed to work pretty coherently pushing our changes for code, revisions, and everything else to Git and things like that.

In terms of users, currently, in my firm, I'm the only user, but the intention is to push it globally for all of our users to be able to use it. 

We would like to be able to support other teams and other departments within the organization. Currently, this is being used only for our credit risk team, but in general, within risk, we have many departments such as operational risk, enterprise risk, market risk, and credit risk. I'm bridging all of them right now. However, with other teams that have expressed an interest, it also will include our settlements team and potentially even our research team and FP&A.

So far, it's been pretty good. I would rate them an eight out of 10. 

People are fairly responsive initially to saying, "Okay, yes, we have this on our radar. Coming back." Sometimes, it might take a little bit longer for some responses, but it's still very good, and the quality is a 10 out of 10.

Positive

At my current firm, we weren't using anything in this team. I just came in, and I knew I wanted to use this product. I had used it quite heavily at my previous firm, and it was just very easy. Even the folks who did not have prior coding experience or data ETL experience could fairly quickly learn its semantics or the ways to work with it. So, I figured that it would be a great product to push forward.

Other teams in my firm were using low-code or no-code solutions, but I just can't stand their interfaces. It's rather limited in terms of even viewing what's on the screen and what you have. I appreciate the way you can debug very quickly within PDI.

It was pretty straightforward for me. I had no problem with configuring it. For my personal use of the product, it took an hour of my time to get it onto my machine. For the Enterprise edition, the deployment is still going on, but it's mainly because we don't have many people on our infrastructure team to help. They have multiple ongoing projects. 

The implementation strategy for my personal use case was fairly straightforward. It involved getting the Community edition and configuring it so that I can set up the pipelines for connecting to my data sources and databases and then output to a file share drive for now. All our databases are fairly read-only on our side. In terms of the implementation strategy for the Enterprise edition, we haven't gotten to the stage of completing it, but it'll work somewhat similarly. It's just that the repositories, instead of them being folder repositories, are going to be database-driven, and any code is going to be pushed to the database repository.

We are not using any integrator or consultant for this. For its deployment and maintenance, we're rather limited in terms of the staff. We have one infrastructure person and me. I'm going to be in charge of maintaining it for the time being until I can increase my team.

When you can get things done much faster and free up people's time, it's a no-brainer.

When I came into the firm, I was using the Community edition, which is the freeware version. Because the Enterprise edition costs something, it has actually increased our costs, but as a whole, in terms of operational ability and time savings for the rest of my team, the output from PDI and everything else has only increased the value of using this product.

The pricing has been pretty good. I'm used to using everything open-source or freeware-based. I understand that organizations need to make sure that the solutions are secure, and that's basically where I hit a roadblock in my current organization. They needed to ensure that we had a license and we had a secure way of accessing it so that no outside parties could get access to our data, but in terms of pricing, considering how much other teams are spending on cloud solutions or even their existing solutions, its price point is pretty good.

At this time, there are no additional costs. We just have the licensing fees.

If you don't have the comfort level for the architectural build-out, then you can definitely opt for the white gloves treatment with an additional cost of about 50,000 to help with the integration and implementation effort of it. We chose not to go that route. Therefore, we're using support for any of the fine-tuning questions about making it highly available and other things.

I have not used Lumada for creating pipelines. I'm using PDI to help with our data pipelines. Similarly, I am not using its ability to develop and deploy data pipeline templates at this time, and I also haven't used it for single end-to-end data management from ingestion to insight.

The biggest lesson that I have learned from using this solution is that the order of operations is critical. Other than that, it has been an absolute treat to use.

I've been espousing this product to everybody. I would rate it a 10 out of 10.

We are a service delivery enterprise, and we have different use cases. We deliver solutions to other enterprises, such as banks. One of the use cases is for real-time analytics of the data we work with. We take CDC data from Oracle Database, and in real-time, we generate a product offer for all the products of a client. All this is in real-time. The client could be at the ATM or maybe at an agency, and they can access the product offer. 

We also use Pentaho within our organization to integrate all the documents and Excel spreadsheets from our consultants and have a dashboard for different hours for different projects.

In terms of version, currently, Pentaho Data Integration is on version 9, but we are using version 8.2. We have all the versions, but we work with the most stable one. 

In terms of deployment, we have two different types of deployments. We have on-prem and private cloud deployments.

I work with a lot of data. We have about 50 terabytes of information, and working with Pentaho Data Integration along with other databases is very fast.

Previously, I had three people to collect all the data and integrate all Excel spreadsheets. To give me a dashboard with the information that I need, it took them a day or two. Now, I can do this work in about 15 minutes.

It enables us to create pipelines with minimal manual coding or custom coding efforts, which is one of its best features. Pentaho is one of the few tools with which you can do anything you can imagine. Our business is changing all the time, and it is best for our business if I can use less time to develop new pipelines.

It provides the ability to develop and deploy data pipeline templates once and reuse them. I use them at least once a day. It makes my daily life easier when it comes to data pipelines.

Previously, I have used other tools such as Integration Services from Microsoft, Data Services for SAP, and Informatica. Pentaho reduces the ETL implementation time by 5% to 50%.

Pentaho from Hitachi is a suite of different tools. Pentaho Data Integration is a part of the suite, and I love the drag-and-drop functionality. It is the best. 

Its drag-and-drop interface lets me and my team implement all the solutions that we need in our company very quickly. It's a very good tool for that.

Their client support is very bad. It should be improved. There is also not much information on Hitachi forums or Hitachi web pages. It is very complicated.

In terms of the flexibility to deploy in any environment, such as on-premise or in the cloud, we can do the cloud deployment only through virtual machines. We might also be able to work on different environments through Docker or Kubernetes, but we don't have an Azure app or an AWS app for easy deployment to the cloud. We can only do it through virtual machines, which is a problem, but we can manage it. We also work with Databricks because it works with Spark. We can work with clustered servers, and we can easily do the deployment in the cloud. With a right-click, we can deploy Databricks through the app on AWS or Azure cloud.

I have been using Pentaho Data Integration for 12 years. The first version that I tested and used was 3.2 in 2010.

Their technical support is not good. I would rate them 2 out of 10 because they don't have good technical skills to solve problems.

Negative

It is very quick and simple. It takes about five minutes.

I have a good knowledge of this solution, and I would highly recommend it to a friend or colleague. 

It provides a single, end-to-end data management experience from ingestion to insights, but we have to create different pipelines to generate the metadata management. It's a little bit laborious to work with Pentaho, but we can do that.

I've heard a lot of people say it's complicated to use, but Pentaho is one of the few tools where you can do anything you can imagine. It is very good and quite simple, but you need to have the right knowledge and the right people to handle the tool. The skills needed to create a business intelligence solution or a data integration solution with Pentaho are problem-solving logic and maybe database knowledge. You can develop new steps, and you can develop new functionality in Pentaho Lumada, but you must have the knowledge of advanced Java programming. Our experience, in general, is very good. 

Overall, I am satisfied with our decision to purchase Hitachi's product services and solutions. My satisfaction level is at an eight out of ten.

I am not much aware of the roadmap of Hitachi Vantara. I don't read much about that.

I would rate this solution an eight out of ten. 

The company where I was working previously was using this product. We were using it for ETL process management. It was like a data flow automatization.

In terms of deployment, we were using an on-premise model because we had sensitive data, and there were some restrictions related to information security.

Our data flow processes became faster with this solution.

It has a really friendly user interface, which is its main feature. The process of automating or combining SQL code with some databases and doing the automation is great and really convenient.

As far as I remember, not all connectors worked very well. They can add more connectors and more drivers to the process to integrate with more flows.

The last time I saw this product, the onboarding instructions were not clear. If the process of onboarding this product is made more clear, it will take the product to the next level. There is a possibility that the onboarding process has already improved, and I haven't seen it. 

I have used this solution for two or three years.

I would rate it an eight out of ten in terms of stability.

We didn't have to scale too much. So, I can't evaluate it properly in terms of scalability.

In terms of its users, only our team was using it. There were approximately 20 users. It was not for the whole company.

We didn't use too much customer support. We were using the open-source resources through Google Search. So, we were just using text search. There were some helpful forums where we were able to find the answers to our questions.

I didn't use any other solution previously. This was the only one.

I wasn't a part of its deployment. In terms of maintenance, as far as I know, it didn't require much maintenance.

We absolutely saw an ROI. It was hard to calculate, but we felt it in terms of
the speed of our processes. After using this product, we could do some of the things much faster than before.

I mostly used the open-source version. I didn't work with a license.

I did not evaluate other options.

I would recommend using this product for data engineering and Extract, Transform, and Load (ETL) processes.

I would rate it an eight out of ten.

My work primarily revolves around data migration and data integration for different products. I have used them in different companies, but for most of our use cases, we use it to integrate all the data that needs to flow into our product. Also, we can have outbound from our product when we need to send to different, various integration points. We use this product extensively to build ETLs for those use cases.

We are developing ETLs for the inbound data into the product as well as outbound to various integration points. Also, we have a number of core ETLs written on this platform to enhance our product.

We have two different modes that we offer: one is on-premises and the other is on the cloud. On the cloud, we have an EC2 instance on AWS, then we have installed that EC2 instance and we call it using the ETL server. We also have another server for the application where the product is installed.

We use version 8.3 in the production environment, but in the dev environment, we use version 9 and onwards.

We have been able to reduce the effort required to build sophisticated ETLs. Also, we now are in the migration phase from an on-prem product to a cloud-native application. 

We use Lumada’s ability to develop and deploy data pipeline templates once and reuse them. This is very important. When the entire pipeline is automated, we do not have any issues in respect to deployment of code or with code working in one environment but not working in another environment. We have saved a lot of time and effort from that perspective because it is easy to build ETL pipelines.

The metadata injection feature is the most valuable because we have used it extensively to build frameworks, where we have used it to dynamically generate code based on different configurations. If you want to make a change at all, you do not need to touch the actual code. You just need to make some configuration changes and the framework will dynamically generate code for that as per your configuration. 

We have a UI where we can create our ETL pipelines as needed, which is a key advantage for us. This is very important because it reduces the time to develop for a given project. When you need to build the whole thing using code, you need to do multiple rounds of testing. Therefore, it helps us to save some effort on the QA side.

Hitachi Vantara's roadmap has a pretty good list of features that they have been releasing with every new version. For instance, in version 9, they have included metadata injection for some of the steps. The most important elements of this roadmap to our organization’s strategy are the data-driven approach that this product is taking and the fact that we have a very low-code platform. Combining these two is what gives us the flexibility to utilize this software to enhance our product.

It could be better integrated with programming languages, like Python and R. Right now, if I want to run a Python code on one of my ETLs, it is a bit difficult to do. It would be great if we have some modules where we could code directly in a Python language. We don't really have a way to run Python code natively. 

I have been working with this tool for five to six years.

They are making it a lot more stable. Earlier, stability used to be an issue when it was not with Hitachi. Now, we don't see those kinds of issues or bugs within the platform because it has become far more stable. Also, we see a lot of new big data features, such as connecting to the cloud.

Lumada is flexible to deploy in any environment, whether on-premises or the cloud, which is very important. When we are processing data in batches on certain days, e.g., at the end of the week or month, we might have more data and need more processing power or RAM. However, most times, there might be very minimal usage of that CPU power. In that way, the solution has helped us to dynamically scale up, then scale down when we see that we have more data that we need to process.

The scalability is another key advantage of this product versus some of the others in the market since we can tweak and modify a number of parameters. We are really impressed with the scalability.

We have close to 80 people who are using this product actively. Their roles go all the way from junior developers to support engineers. We also have people who have very little coding knowledge and are more into the management side of things utilizing this tool.

I haven't been part of any technical support discussions with Hitachi.

We are very satisfied with our decision to purchase Hitachi's product. Previously, we were using another ETL service that had a number of limitations. It was not a modern ETL service at all. For anything, we had to rely on another third-party software. Then, with Hitachi Lumada, we don't have to do that. In that way, we are really satisfied with the orchestration or cloud-native steps that they offer. We are really happy on those fronts.

We were using something called Actian Services, which had less features and it ended up costing more than the enterprise edition of Pentaho.

We could not do a number of things on Actian. For instance, we were unable to call other APIs or connect to an S3 bucket. It was not a very modern solution. Whereas, with Pentaho, we could do all these things as well as have great marketplaces where we could find various modules and third-party plugins. Those features were simply not there in the other tool.

The initial setup was pretty straightforward. 

We did not have any issues configuring it, even in my local machine. For the enterprise edition, we have a separate infrastructure team doing that. However, for at least the community edition, the deployment is pretty straightforward.

We have seen at least 30% savings in terms of effort. That has helped us to price our service and products more aggressively in the market, helping us to win more clients.

It has reduced our ETL development time. Per project, it has reduced by around 30% to 35%.

We can price more aggressively. We were actually able to win projects because we had great reusability of ETLs. A code that was used for one client can be reused with very minimal changes. We didn't have any upfront cost for kick-starting projects using the Community edition. It is only the Enterprise edition that has a cost. 

For most development tasks, the Enterprise edition should be sufficient. It depends on the type of support that you require for your production environment.

We did evaluate SSIS since our database is based on Microsoft SQL server. SSIS comes with any purchase of an SQL Server license. However, even with SSIS, there were some limitations. For example, if you want to build a package and reuse it, SSIS doesn't provide the same kinds of abilities that Pentaho does. The amount of reusability reduces when we try to build the same thing using SSIS. Whereas, in Pentaho, we could literally reuse the same code by using some of its features.

SSIS comes with the SQL Server and is easier to maintain, given that there are far more people who would have knowledge of SSIS. However, if I want to do a PCP encryption or make an API connection, it is difficult. To create a reusable package is not that easy, which would be the con for SSIS. 

The query performance depends on the database. It is more likely to be good if you have a good database server with all the indexes and bells and whistles of a database. However, from a data integration tool perspective, I am not seeing any issues with respect to query performance.

We do not build visualization features that much with Hitachi. For the reporting purposes, we have been using one of the tools from the product, then prepare the data accordingly. 

We use this for all the projects that we are currently running. Going forward, we will be sticking only to using this ETL tool.

We haven't had any roadblocks using Lumada Data Integration.

On a scale of one to 10, I would recommend Hitachi Vantara to a friend or colleague as a nine.

If you need to build ETLs quickly in a low-code environment, where you don't want to spend a lot of time on the development side of things but it is a little difficult to find resources, then train them in this product. It is always worth that effort because it ends up saving a lot of time and resources on the development side of projects.

Overall, I would rate the product as a nine out of 10.

My primary use case is to provide integration with my source systems, such as ERP systems and SAP systems, and web-based systems, having them primarily integrate with my data warehouse. For this process, I use ETL to treat and gather all the information from my first system, then consolidate it in my data warehouse.

We needed to gather data from many servers at my company. We had probably 10 or 12 equivalent databases spread around the world, i.e., Brazil, Paraguay, or Chile, and had an instance in each country. So, this server is Microsoft SQL Server-based. We are using Lumada to get the data from these international databases. We can parallelize the extraction from various servers at the same time because we have the same structure, schemas, and tables in each of these SQL Server-based servers. This provides a good value for us, as we can extract data at the same time in parallel, which accelerates our extraction.

In one integration process, I can retrieve data from 10 or 12 servers at the same time in one transformation. In the past, using SQL Server or other manual tools, we needed to have 10 or 12 different processes, one per server. Using Lumada in parallel accelerates our extraction. The tools that Lumada provides enable us to transform the data during this process, integrating the data in our data warehouse with good performance. 

Because Lumada uses Java virtual machines, we can deploy and operate in whatever operational system that we want. We can deploy on Linux, even when we had a Linux version from Lumada and a Windows version from Lumada.

It is simple to deploy my ETLs because Lumada has the Pentaho Server version. I installed the desktop version so we can deploy our transformations in the repository. We install our own Lumada on a server, then we have a web interface to schedule our ETLs. We are also able to reschedule our ETLs. We can schedule the hour that we want to run our ETL processes and transformations. We can schedule how many times we want to process the data. We can save all our transformations in a repository located in a Pentaho Server. Since we have a repository, we can save many versions of our transformation, such as 1.0, 1.1, and 1.2, in the repository. I can save four or five versions of a transformation. I can ask Lumada to run only the last version that I saved in the database. 

Lumada offers a web interface to follow these transformations. We can check the logs to see if the transformations were successfully completed, we had a network query, or some database log issues. Using Lumada, there is a feature where we can get logs at the execution time. We can also be notified by email if transformations occurred successfully or failed. We have a file for each process that we schedule on Pentaho Server.

The area where Lumada has helped us is in the commercial area. There are many extractions to compose reports about our sales team performance and production steps. Since we are using Lumada to gather data from each industry in each country. We can get data from Argentina, Chile, Brazil, and Colombia at the same time. We can then concentrate and consolidate it in only one place, like our data warehouse. This improves our production performance and need for information about the industry, production data, and commercial data.

The features that I use the most are Microsoft Excel table input, S3 CSV Input, and CSV input. Today, the features that are more valuable to me are the table input, then the CSV input. These both are very important. We extract data from the table system for our transactional databases, which are commonly used. We also use the CSV input to get data from AWS S3 and our data lake.

In Lumada, we can parallelize the steps. The performance to query the databases for me is good, especially for transactional databases. Because Lumada uses Java, we can adjust the amount of memory that we want to use to do transformations. So, it is accessible. It's possible to set up the amount of memory that we want to use in the Java VM, which is good. Therefore, Lumada is good, especially with transactional database extraction. It has good performance, not higher performance, but good performance as we query data, and it is possible to parallelize the query. For example, if we have three or four servers to get the data, then we can retrieve the data at the same time, in parallel, in these databases. This is good because we don't need to wait while one of the extractions finishes. 

Using Lumada, we don't need to do many manual transformations because we have a native company for many of our transformations. Thus, Lumada is a low-code tool to gather data from SQL, Python, or other transformation tools.

Lumada could have more native connectors with other vendors, such as Google BigQuery, Microsoft OneDrive, Jira systems, and Facebook or Instagram. We would like to gather data from modern platforms using Lumada, which is a better approach. As a comparison, if you open Power BI to retrieve data, then you can get data from many vendors with cloud-native connectors, such as Azure, AWS, Google BigQuery, and Athena Redshift. Lumada should have more native connectors to help us and facilitate our job in gathering information from these new modern infrastructures and tools.

I have been using Lumada Data Integration for at least four years. I started using it in 2018.

Because we are using the free version of Lumada, we have used only the support on the communities and forums on the Internet. 

Lumada does have a paid version, where Hitachi support is specialized in Lumada support. 

It is simple to deploy Lumada because we can deploy our transformation in three to five simple steps, saving our transformation in a repository. 

I open the Pentaho Server web-based version, then I find the transformation that I deployed. I can schedule this transformation at the hour or recurrence in which I want to run the transformation. It is easy. Because at the end of the process, I can save my transformation and Lumada generates the XML file. We can send this XML file to any user of Lumada, who can open up this model and get the transformation that I developed. As a deployment process, it is straightforward, simple, and not complex.

Using Lumada compared to using SQL manually, ETL development time is half the time it took using a basic manual transformation.

There are more types of connectors, but you need to pay. 

You need to go through the paid version to have Hitachi Lumada specialized support. However, if you are using the free version, then you will have only the community support. You will depend on the releases from Hitachi to solve some problem or questions that you have, such as bug fixes. You will need to wait for the newest versions or releases to solve these types of problems.

I also use Talend Data Integration. For me, Lumada is straightforward and makes it simpler to have transformations as drag and drops. Comparing Talend and Lumada, I think Lumada is easier to use, more than Talend. The comprehension needed for these tools is less with Lumada with than Talend. I can learn Lumada in a day and proceed with my transformations, using some tutorials, since Lumada is easier to use. Whereas, Talend is a more complex solution with more complex transformations.

In Talend's open version, i.e., free version, you won't have a Talend server to deploy models. Thus, you deploy Talend models on the server. If you want to schedule some transformation, then you need to use the operational system where you have infrastructure to run transformations and deploy them. For example, in Talend, we deployed a data model in Talend, but we needed to use Windows Scheduler to also schedule the packets in Talend to process the data in the free version of Talend. Whereas, in the free version of Lumada, we already had it based on the web server. Therefore, we can run our transformations and deploy them on the server. We can schedule in a web interface, which guides us with scheduling the data and checking our logs to see how many transformations we have at a time. This is the biggest difference between Talend and Lumada.

I don't use many templates. I use the solution based on a case-by-case basis.

Considering that Lumada is a free tool, I would rate it as nine out of 10 for the free version.

We use Pentaho for small ETL integration jobs and cross-storage analytics. It's nothing too major. We have it deployed on-premise, and we are still on the free version of the product.

In our case, processing takes place on the virtual machine where we installed Pentaho. We can ingest data from different on-premises and cloud locations. We still don't carry out the data processing phase inside a different environment from where the VM is running.

At the start of my team's journey at the company, it was difficult to do cross-platform storage analytics. That means ingesting data from different analytics sources inside a single storage machine and building out KPIs and some other analytics. 

Pentaho was a good start because we can create different connections and import data. We can then do some global queries on that data from various sources. We've been able to replace some of our other data tools like Talend for our managing data warehouse workflow. Later, we adopted some other cloud technologies, so we don't primarily use Pentaho for those use cases anymore. 

Pentaho is flexible with a drag-and-drop interface that makes it easier to use than some other ETL products. For example, the full stack we are using in AWS does not have drag-and-drop functionality. Pentaho was a good option at the start of this journey.

We can schedule job execution in the BA Server, which is the front-end product we're using right now. That scheduling interface is nice.

It's difficult to use custom code. Implementing a pipeline with pre-built blocks is straightforward, but it's harder to insert custom code inside the pre-built blocks. The web interface is rusty, and the biggest problem with Pentaho is debugging and troubleshooting. It isn't easy to build the pipeline incrementally. At least in our case, it's hard to find a way to execute step by step in the debugging mode.

Repository management is also a shortcoming, but I'm not sure if that's just a limitation of the free version. I'm not sure if Pentaho can use an external repository. It's a flat-file repository inside a virtual machine. Back in the day, we would want to deploy this repository on a database.

Pentaho's data management covers ingestion and insights but I'm not sure if it's end-to-end management—at least not in the free version we are using—because some of the intermediate steps are missing, like data cataloging and data governance features. This is the weak spot of our Pentaho version.

We implemented Hitachi Pentaho some time ago. We have been using it for around five or six years. I was using the product at the time, but now I am the head of the data engineering team, so I don't use it anymore but I know Pentaho's strengths and weaknesses.

Pentaho is relatively stable, but I average about one failed job every month. 

I rate Pentaho six out of 10 for scalability. The scalability depends on how you deploy it. In our case, the on-premise virtual machine is relatively small and doesn't have a lot of resources. That is why Pentaho does not handle big datasets well in our case. 

I'm also unsure if we can deploy Pentaho in the cloud. So when you're not dealing with the cloud, scalability is always limited. We cannot indefinitely pump resources into a virtual machine.

Currently, we have five or six active workflows running each night. Some of them are ingesting data from ADU. Others take data from AWS Redshift or on-premise Oracle. In terms of people, three other people on the data engineering team and I are actively using Pentaho.

We used Talend, which is a Java-based solution and is made for people with proficiency in Java. The entire analytics ecosystem is transitioning to more flexible runtimes, including Python and other languages. Java was not ideal for our data analytics journey.

Right now, we are using NiFi, a tool in the cloud ecosystem that has a similar drag-and-drop interface, but it's embedded in the ADU framework. We're also using another drag-and-drop tool on AWS, but not AWS Glue Studio. 

We've seen a 50 percent reduction in our ETL development time using the free version of Pentaho. That saves about 1,000 euros per week, so at least 50,000 euros annually. 

I rate Pentaho eight out of 10. It's a perfect pick for data teams that are getting started and more business-oriented data teams. It's good for a data analyst who isn't so tech-savvy. It is flexible and easy to use.