Humio

an excellent APM tool for large data applications

Commentaires : I am very happy with the product. since it can handle very large data with an exceptional processing speed.
It has many advanced features. due to thee many features, in initial felt like GUI is very complex. but later realized GUI is very easy.

Avantages :

Data ingestion speed is very fast It can handle very huge data can customize the dashboards in simple real time streaming help from community groups

Inconvénients :

In initial, the GUI is very confused, due to many features, but later its very easy to use. sometimes GUI takes long time to load the data

Great tool that makes it easy to analyze log files from your servers or cloud. It's blazing fast!

Commentaires : In the beginning we tried to analyze our log files with linux commands like grep / tail and created a lot of custom scripts for doing so. This was cumbersome and error-prone.
We were thinking about setting up elasticsearch, but ended up using Humio. It was the right choice for us.
It is easy to get started. We use filebeat for ingesting most of our data, but we also make use of their ingest api for some odd cases. When the data is uploaded it is actually fun to analyze and work with the data as Humio is blazing fast and allows us to visualise our logs.

Avantages :

I really like their intuitive query language that works much like pipes in linux. Eg. the query "error | count()" counts all the log lines that contains the string "error". More complex queries can be constructed by chaining and combining multiple built in functions for grouping events and parsing raw data in the events. Queries can easily be converted into widgets on a dashboard. We use their dashboarding feature for monitoring server and network health status. We currently have ingested more than 100GB of data and most queries only takes a few seconds to complete! It's way faster than searching with ack and even faster than searching with ag/"the silver surfer" on my desktop with SSD.

Inconvénients :

Their pricing model is not well fitted for startups. If you need more than 7 days of retention or more than one user, then you need to pay.

Capable of showing you the full overview of your systems, that you never thought were possible.

Avantages :

You can create your very own custom dashboards, for both internal and external use. Using the functions in queries, you can customize a large variety, of useful overviews, which enables you to react faster and take action on events in your own systems. It gives a certain comfort, and preparedness, knowing that you can rely on the data from the logs.

Inconvénients :

The server nodes is occasionally down, for less or longer time. Resulting in time where real-time overview is "stopped" - but no logs are lost.

Raw log analysis power at your fingertips

Avantages :

Setting up log collection with Humio is pretty easy, and once your log data is shipping to Humio, it's just a matter of digging out answers to all your questions using Humio's query language. Humio is FAST, and it's an amazing feeling to be able to get such instant results even on advanced queries in a large amount of log data. Once you've experienced Humio's speed, you don't want to go back to whatever you used before. There is a good selection of graph tools available, and you can create nice dashboards with live data. There's also support for alerts, but I haven't used those yet, so I don't know how well they work. Judging from the rest of the system, it's probably well-thought-out.

Inconvénients :

The GUI is not super pretty and can be a bit confusing at times, e.g. when editing graphs on a dashboard. However, once you get to know how to navigate it, it's fast and functional.

Powerful tool to deep dive into how my application actually works and not just how I think it works

Commentaires : I know a lot more about my application than I did before I started using Humio. So now there are fewer errors in my application - and when there are errors they are fixed a lot faster.

Avantages :

It is so incredibly fast that you can try any query that may or may not give a useful result without feeling you are wasting time. If you simply remember to log anything that may become useful information sometime in the future you can easily find it again. Additionally, Humio approach that everything is searched and you can use regular expressions over the entire event makes it incredibly powerful.

Inconvénients :

The support for structured logging is not perfect yet, so the first thing you always do is to write | table([@timestamp, messageTemplate])