Apache Tomcat VS zio-http

Manual instrumentation allows you to define your Spans within the code itself rather than relying on automatic instrumentation finding the entry point for a trace. Manual instrumentation is especially helpful for applications that don’t use an application server such as Tomcat, JBoss, or Jetty.

99% is a huge exaggeration. Two essential deployment tools off the top of my head: https://tomcat.apache.org/ https://docs.jboss.org/author/display/AS71/Developer%20Guide.html

To download, enter the Apache website in the address bar: http://tomcat.apache.org/

Do we still enjoy it? We are running many Vaadin apps in production since that first one. If there are not any specific requirements we use a “modular monolith” concept, which fits our stack best. We pack applications as WAR and deploy them under Apache Tomcat. And yes, we enjoy the development process. It’s very straightforward and Vaadin and SpringBoot fit together well.

JasperReports Server Community requires a Java application server and a database to create a repository in order to work properly. After downloading JRS, the installation process can install Tomcat server and PostgreSQL database automatically for us and the services will run depending on the Jasper server. It's also possible to connect JRS to services already installed on the server. Moreover, while the free version supports the MySQL server, with the paid version you can also use commercial databases, for example, ORACLE Database, as a data repository.

Regarding the venerable Tomcat, they [somewhat] recently added support for Unix domain sockets.

* https://github.com/apache/tomcat/pull/402

* https://github.com/apache/tomcat/pull/532

We fronted the server with haproxy LTS. Our initial testing showed roughly and order of magnitude [10x] increase in the number of requests the server could handle.

It's not completely plug-and-play; we still had a write a custom valve to set the request remote ip address and some other TCPish stuff, but nevertheless the capacity far outstripped our need for the technology.

Official Website :- https://tomcat.apache.org/

Don't use an installed copy of Tomcat. The layout can be different than expected and permission problems can appear at the worst time. For one, it needs to be able to write to that conf directory. Download a non-platform-specific "core" zip file from tomcat.apache.org instead.

Anytime you use dependencies you make use of code for a purpose that was not originally intended. The apachee web server was named after a patchwork quilt. Many contributors had useful code from commercial projects they contributed to the cause. In networking protocols packets are discarded after a set number of hops, or an expiration date so the internet does not have zombie packets wondering around. The tcp/ip traceroute command uses TTL to find all the collision domains in a request.

If you plan on a full enterprise solution apache server has built in roles, users and corporate tools. It integrates well with Java. Express is much easier in my opinion. You may use xml instead of json. Other than that the front end and backend don't matter. Since ERP is what you are dealing with maybe that is the way to go?

For JNI, make sure you have `-H:+JNI` in your native image options. Be aware that Netty is a bit of a monster to include in a native image application (I just recently added it a Netty dependency and it completely broke my project and took a good day to figure out). Mine was through zio-http and I was able to piece together some working options from github issues: https://github.com/zio/zio-http/issues/20.

What's the hardware being used for your test? I get 55k RPS with a basic 200 responder with zio-http[0] (which uses Netty) on my i5-6600K, and over 20k RPS for an e2e POST endpoint that does write batching to postgres (committing the insert before responding with the db generated id). Postgres, client (vegeta[1]), and the app all on the same machine. I think that was with keep-alive, I think like 256 clients for the basic responder and 1024 for the one that writes to the db. There's a recently merged PR for zio-http that does 1M req/s on whatever machine they test on[2] so Netty can absolutely scale to high RPS.

[0] https://github.com/zio/zio-http

[1] https://github.com/tsenart/vegeta

[2] https://github.com/zio/zio-http/pull/1659

Yep! https://github.com/zio/zio-http

I was looking at HTTP frameworks for Scala. I found zio-http, https://zio.github.io/zio-http/. Has anyone tried this framework?

https://github.com/dream11/zio-http very simple and super performant, dead easy to abstract over routes to make a router

I'm starting a new project, essentially a standard web app (SPA + rest api). I'd like to stay as functional as possible. I'm debating between http4s + cats effect, http4s + zio, and zio-http (https://github.com/dream11/zio-http). I'm having a hard time figuring out which stack has the strongest (safest?) future especially with scala 3 out. Cats Effect 3 looks amazing, but so does the zio ecosystem. Thanks for any opinions/advice!