Splunk Pros and Cons

It is very simple to tweak or write a small piece of glue code to go ahead and create a new dashboard for a business unit to make near real-time decisions to focus more on other geographies when launching the product.

On the cloud, we are pushing through less than half a petabyte of data. So far, it has been fairly stable because it runs on all the underlying AWS infrastructures.

The ability to view all of these different logs, then drilling down into specific times or into specific data sources, has proved to be the greatest aspect in decreasing our troubleshooting overhead time.

We did not encounter any issues with scalability. It is almost seamless to add new index (storage) or search (used to analyze the data) nodes to the cluster.

Its usability is the best part. It is easy for our developers to use if they want to search their logs, etc.

The dashboards are the most valuable feature. We like the ability to drill in and see what queries are under the dashboard, build new visualizations, edit the querying, and see the reports.

We were able to create a catalog of dashboards and have a holistic view at all levels. We could understand our business much better. Real-time errors, which were buried in emails before now, surfaced up on dashboards.

We have a one stop dashboard for health of some of our services where you can click in and it takes you to other dashboards that have custom near real-time metrics that show the application's health.

The technical support has been very good. They are very responsive and have been helpful.

You can run reports against multiple devices at the same time. You are able to troubleshoot a single application on a thousand servers. You can do this with a single query, since it is very easy to do.

With good domain knowledge, one can build almost anything. If you throw in Alert Manager or an integration with ServiceNow. Then, you have your own SIEM

Our clients are easily able to modify and evolve their implementations.

The initial setup is simple, not very complex. Initial deployment takes around 10 to 15 minutes to set up the entire base for Splunk including all three tiers.

It's extremely scalable. It's a very robust solution and certainly has the capability of handling far bigger data requirements than a lot of the other tools. Generally what ends up happening with me is that my clients tend, for the most part, to be mid-tier organizations where the cost of that solutions would be accompanying requirements for people just becomes way too prohibitive. Especially considering the model that they use for costing, which is based on the volume of data. Of course, they're going to put everything including the Coke machine as the ability to collect data off of it, because of course the more they can put through the tool the more money they make.

It has reduced the time to resolution, time to investigate, and time to troubleshoot for debugging issues.

Out-of-the-box, it seems very powerful.

For on-premise, it's more about optimization. With such a heavy byte scale of data that we are operating on, the search for disparate data sometimes takes about a minute. This is understandable considering the amount of data that we are pumping into it. The only optimization that I recommend is better sharding, when it comes to Splunk, so that data retrieval can be faster.

I feel as though a major focus of upcoming releases should be set on Machine Learning, Predictive Analytics, and I would enjoy to see more security focused add-ons and apps developed by the vendor.

A problem that we had recently had was we licensed it based on how much data you upload to them every day. Something changed in one our applications, and it started generating three to four times as many logs and. So now, we are trying to assemble something with parts of the Splunk API to warn ourselves, then turn it off and throttle it back more. However it would be better if they had something systematically built into the product that if you're getting close to your license, then to shut things down.

We do have to educate developers on how to not blow it up. It is a little to easy to write an expensive query and overly stress the system. This could be improved.

I would like additional features in different programming models with the support for writing queries in SQL or other languages, such as C#, Java, or some other type of query definitions.

The historical data extraction needs improvement. I would like the capability of taking data and having it trend longer.

When you get into large amounts of data, Splunk can get pretty slow. This is the same on-premise or AWS, it doesn't matter. The way that they handle large data sets could be improved.

I would like to see an updated dashboard. The dashboard is a little out-of-date. It could be made prettier.

It needs a better way to export dynamic views without requiring a ton of code and user/pw.

It needs integration with a configuration management solution.

It needs integration with a configuration management solution.

Splunk does not build apps. They only go back and validate the apps that somebody has already built. They should have remote consulting support. They have a wonderful solution. They have 24/7 security. Nobody needs to depend on any third party and will therefore just buy Splunk on the cloud.

The tool itself is very difficult to configure. It's great for its number of inputs, for the different types of systems devices, and things that it could collect information from. To actually make good use of it, you need a fairly dedicated team of people that have some reasonably good programming or modeling skills to be able to do the things that you need to do with it. Whereas a lot of the other tools are better packaged for that, and so require a lot less training and a lot less dedication.

My company could benefit from doing more Splunk training with Splunk consultants teaching us how to use it.