Heimdall Data Reviews
by Heimdall Data
Reviewed on 12/02/2019
Great product with awesome support
Comments: The problem we are trying to solve is providing means for intelligent automatic failover for database server clusters. At the same time it is great to have the performance enhancing features such as intelligent read/write splits caching. Support has been oustanding. Heimdall data, like the solutions before it was not 100% trouble free to setup. The difference with Heimdall though was the quality and responsiveness of the support. I got support turnarounds in hours or even less with at least info on what the problem is if not communication that they understood the issue and will have a fix for it soon. Compare to a (worst case) scenario with a previous vendor where it literally took more than two years to get resolution on one SQL distribution problem.
Pros: Heimdall Data is a database load distribution, caching, failover and high availability solution. I have worked with two other companies that offered a subset of this functionality over the past 7 years. There are very few solutions on the market because this is a very complex problem to tackle. What I liked most about the software was probably it's ability to achieve all of these functions within a single product. It's caching ability is unlike any other product I have seen. It's quite good at load distribution and automatic failover. It also has a decentralized model for the traffic routing, which can be good especially in cases where there is concern about a centralized bottleneck in the traffic flow (and/or CPU requirements for SQL routing)
Cons: Ease of configuration could be improved, manually setting up the rules to do read/write split for example (while easy to do if you have the documentation) could be better handled with simple toggles for most use cases(example). The decentralized model has advantages over a centralized model for scalability especially for more distributed applications (e.g. micro services). The previous solutions I used were centralized only. The drawback to decentralized is it can consume quite a bit more resources depending on the number of app servers that you have since you are literally running multiple copies of the software one per application server.
Reviewed on 07/02/2017
Insightful visibility into Wirespeed Database interactions
Comments: It was a pleasure to find Heimdall Data and work with their team on several projects. The Heimdall Data Access Platform was the only solution I found for monitoring and evaluating database traffic that was occurring in volumes of 1GB per five minute period. Through use of this tool, I found a vendors defect and pointed to previously unknown and unquantified performance degradations that were periodic. I would recommend this for any database-dependent application because of this visibility. Heimdall Data through optimizations can also remedy peculiarities of a production application or CoS software package that a vendor may be slow to patch. My initial PoC with Heimdall found and documented within a week, persistent and elusive application issues that I had spent months collecting packets for analysis in Riverbed tools. With the packet captures the volume of data was too large and made it impossible to perform troubleshooting of sporadic events. The aggravation and hassle of packet analysis was quite frustrating, I literally lost whole weekends fighting to import packets with out Riverbed crashing. In contrast, The Heimdall Data install was painless on the App server, requiring only a simple configuration and was collecting data immediately on the wire. There was no modification of the Database server required. A good point to note as many DBAs are quite reluctant to install agents on DB servers as a matter of course. Within the week, in consultation with Heimdall Data the anomalies and performance degradations were pinpointed. It was a pleasure working with the whole team and their insight and ability to provide assistance pushed the PoC through to a documented success. I would recommend Heimdall Data to anyone troubleshooting sporadic and performance issues - the icing on the cake is that Heimdall Data can optimize the database traffic as well, and this is a serious issue to consider when dealing with closed source or complex schema. I believe that where Heimdall Data shines is the support they perform in a PoC the expertise they bring to the table is phenomenal and this is why I must give the tool four stars. I think the customer service and rapid turn around that I saw in the PoC demonstrated significant time saved. While you may be able to get up to speed in a year and be prepared to deploy future applications on your own, I would ask for a Professional Services quote first as they are quite efficient and can save you time.
Ability to Optimize DB Traffic
Cons: Moderate Learning Curve during initial deployment.
Reviewed on 15/01/2019
Heimdall for Improved App Speed
Comments: We implemented Heimdall primarily as a caching layer in front of both a standard PostgreSQL datastore as well as an Apache HAWQ datastore. It was very easy to get set up and get started. It was reasonably well integrated with AWS which made the initial configuration much easier. This was a very easy way for us to get started testing Heimdall on something more than a local database.
Pros: Heimdall is easy to set up and get running. Within minutes we had it up and we were testing it with various aspects of our application. The support team was available to us through the entire process, helping us work through what is a unique use case related to how we are using the caching. The support engineer was responsive via email and phone, and worked through various scenarios with us. Super easy install on AWS, which was important to us at the moment. However, installation on Google Cloud is a breeze as well. Heimdall is very flexible. We don't send all of our writes through Heimdall in certain situations, so we had to come up with a more creative way to invalidate the cache. We were able to do that with API calls to Heimdall one one case, and using ruels in another.
Cons: The documentation is largely very good, but there were a few places where we would get confused and need to ask for clarification. It was difficult for us to see the invalidation occurring through the API unless we had the logging turned extremely high, which in turn made it difficult to find the invalidation occurrences That said, we were able to determine is was working with some log spelunking and good testing.
Reviewed on 27/12/2016
Super impressed with ease of implementation
Comments: I was prototyping Heimdall Data's solution with a local application. Primarily to learn it, with the plan of implementing in a production cloud application in the near future. First and foremost, I'm impressed with the ease of installation. I had it up and running my application (python) and a MySQL database. It probably took all of 10 minutes. Once it was up and running, and I found how to configure the 'analyze' component, I literally had a one click cache option. Ok, maybe it was 2-3 clicks to setup, enable and then monitor the speed improvements of my application. The built-in charts/graphs show cache hits, query performance etc. Very slick to see the improvements. I know there are other features, that I just haven't gotten to yet. I'm very interested in the redundant/failover use case. I'll be doing a blog/write up @ http://fauie.com in the super near future about my findings.
Ease of installation.
Performance improvements to my application
Visuals to help see that it's working
by Heimdall Data on 27/12/2016
Glad to hear that you are happy with our "All things SQL" solution. Product feedback is always welcomed.
Reviewed on 06/03/2019
Great product that delivers on what it claims
Comments: Heimdall gave us instant insight into DB issues that had plagued us for months. Armed with data we were able to work with engineering teams to resolve the issues. With the database issues behind us we were able to start using the caching and traffic routing functions which reduced load on databases tremendously. Heimdall saved on the cost of larger databases many times over.
Heimdall is incredibly easy to install compared to other solutions. Given it is a simple driver replacement we were up and running in minutes.
Key functionality + selling points were the following:
Insight into database traffic
Real time query optimization
Database traffic routing
Cons: The decentralized nature of Heimdall can require additional overhead but in turn offers far greater resiliency than its centralized competitors.
Reviewed on 06/11/2017
Heimdall Data's SQL works as advertised quickly providing SQL acceleration on our app.
Comments: Performance plus insight into how to get even more performance. This has allowed us to better satisfy our app customers while cutting the cost of resources.
Pros: Heimdall Data's SQL caching software was easy to install and provided an immediate increase in performance. Their analytic tools gave us insight what was really happening with our queries allowing us to optimize our code. Once optimized, not only was the raw SQL faster but the number of cache hits increased giving us even more performance. If you are serious about performance, you need to start with Heimdall.
Cons: Heimdall is not magic. As I mentioned, part of the problem was with the way that our SQL queries were being made. Heimdall did the caching immediately, but we needed to then use the analytics tools to figure out how to further optimize our code. Once we did that, our cache hits went up and we got even more performance.
Reviewed on 08/05/2018
I am impressed with the simplicity and ease of the implementation process
Comments: In a shot time, your installation will be complete.
Pros: Heimdall Data Logs gives all the data access clearly indicating Who, and what was logged, and the the time. The installation was also faster than we had expected. This is a great product
Cons: It took me longer than my other colleagues to fully get moving, but this could be probably because I was taking my time