Show HN: DevRaven – Monitoring for Developers | Hacker News Reader

Show HN: DevRaven – Monitoring for Developers

devraven.io

46 pointskc1034 comments

Hi HN! I am Krishna Thota, founder of DevRaven. DevRaven is a monitoring platform for Developers.

DevRaven enables engineering teams or individual developers to setup active monitoring for their services/applications and get alerted when things don't work as expected

Today's launch makes available the following features:

API Monitoring - Monitor your HTTP end points and perform no-code or scripted assertions.

Synthetic Monitoring - Execute browser based end-to-end tests using Playwright framework. No setup required.

SSL Monitoring - Monitor SSL certificates for your end points and get alerted before they expire.

Web Page Monitoring - Run continuous Lighthouse audits on your web pages to ensure best performance, SEO.

Welcome any feedback, questions or suggestions.

Open articleView on HN

This is very timely for something I've been wanting to do! Giving it a try now.

Edit: tried it. I like it, I hadn't used playwright before but it was fairly straightforward to set up the test I wanted. Since you requested feedback, some thoughts:

- I started by writing and testing my check locally, and then copied it over into a “synthetic test”. It would be cool if that were more seamless, e.g. if all of my tests were configured through a git repo and DevRaven had a GitHub webhook that was notified when the config changed.

- I had to double-escape some backslashes in my code for it to be accepted, you might be eating the escapes during some string processing?

- It would be cool to see timings of tests (i.e. the time between individual asserts as reported by playwright, but even the overall time would be a useful metric).

- This is probably just my inexperience with playwright, but it would be nice to have an example that used playwright's expect function instead of chai.

- TypeScript support would also be nice.

Thanks Paul for the feedback.

- ya, great idea to be able to pull the changes via git.

- I will look into the issue. edit: confirmed a bug.

- The overall time is captured and is shown in the logs. The dashboard view will be available very soon showing the execution time.

- I will add an example to the recipes for expect as well.

- Agreed on TypeScript support. I will check the feasibility to support on current stack.

update: double-escape issue is now fixed.

That was fast! Cheers.

Hi Krishna, congrats on the launch!

I really like that you have the 'try with no credit card' offering, I wish more companies would do this.

One question I'd have for you is how _your_ services are hosted. If my service is hosted on AWS us-east-1 and so is my monitoring I'll be worried that I'm not going to get alerted when AWS itself has problems.

It seems to me that you're entering into a crowded market. I can think of a dozen or more companies with similar offering. How do you plan on differentiating yourself? Who is the target user for your product and why would they select it instead of a competitor?

Thanks!

Our service is hosted in GCP today. Yes, it's totally possible that alerts may not be sent out or delayed if the regions where we hosted our services are down. I will definitely look into how we can mitigate such issues.

It is a crowded space. But the market opportunity is also huge. I think pretty much any business is a software business. Our main idea is to expand the suite to various other areas of monitoring and be the single platform for monitoring anything.

What differentiate you from the number of existing monitoring platforms? Genuinely curious as it is a crowded space with a lot of established players.

Definitely the space is crowded. But there are wide spectrum of maturity of engineering teams and how much they monitor.

I think the features we have today are the baseline for any monitoring solution. But the idea is to expand and build a single platform to monitor anything that developers would like to monitor.

> a single platform to monitor anything that developers would like to monitor

a jack of all trades, master of none

IMO, nothing wrong with offering core functionality for various types of monitors that works for majority cases.

please think for a moment what you’ve just said

you said you could offer me a tool that does many different things, but nothing specific

but, as a potential customer i would be looking for a tool that could solve my specific problem

now think about that…

Personally speaking, if I have a problem so specific that a generalized tool can't solve it, I am going to create the solution to that problem myself.

If all my problems are so specific they require this treatment then I have too many unicorns and not enough horses.

Thanks for the feedback. I will definitely keep it mind as I navigate through the startup journey.

FWIW: I spent ~10 years in the monitoring / performance space, and your approach in building and marketing a generalist tool is not unreasonable. In fact, more often than not when I saw technology-specific tools, they got pigeon-holed and could not move beyond their original market.

This is an area that needs to be fleshed out, so far the best solution I've found is https://checklyhq.com/

Playwright + other monitoring checks as a service is amazing.

You do have to make your tests pretty resilient to prevent false positives through (retry on network issues etc.)

The Playwright test framework helps with that.

Playwright is amazing. Tests are reliable compared to other frameworks and work on all browsers. That's the reason we use Playwright for running our Synthetic tests.

Checkly is in the market for the last 5+ years and they are definitely feature rich compared to DevRaven. But functionally pretty similar.

this looks nice. generous free tier! seconding the idea that it would be nice if gitops (flat config files in git) was a first class interface to configure everything. that would be a differentiator for me.

Thanks.

Completely agree with your feedback on git being the config source for scripts.

i'd consider going a step further and making the web ui completely optional for configuration after the billing is set up. i know that sounds extreme, but personally that is something i would really, really like. maybe a small cli tool that gives me a template config file that i can then edit and and then run again manually or via ci to apply my changes.

that may be extreme and reflective of my personal preferences, so grains of salt are warranted. :)

:-) I will definitely spike it asap and make it as simple as possible to setup and run via a git repo.

Hi Krishna, Great product you have put there.

What is the long term plan wrt pricing, what should I expect if I sign up?

Like the clarity of the home page, well done and good luck!

Thanks!

The free SKU allow you to start setting up and try out monitoring. You can change to paid SKU when you are ready to run more than 10000 API checks/month or 2500 synthetic checks/months.

We do not have per seat pricing. It's purely usage based and you can purchase additional checks as you go.

Nice work!

Any plans to support cron monitoring / time of day monitoring? Something akin to cronitor?

Thanks!

The idea is to expand the feature set depending on user feedback. Cron monitor is on my radar.

Is there something you don't like in Cronitor?

Never understand why these kind of services only support PagerDuty. What about Opsgenie

hi, it's just a matter of prioritization and time given the resources. I will look into adding Opsgenie support very soon.

I use Checkly. Any idea how this compares?

Checkly is also in the same space and has been in the market for 5+ years. So, it's definitely feature rich compared to DevRaven.

But the core functionality is same for both the products i.e. the ability to run browser checks or API checks at scheduled intervals. We will mature and add more features as we go.

why would i use your proprietary product when there are (lots!) of open-source alternatives available?

Whether to use a SaaS service or build something in-house is a choice for your engineering organization.

But in most cases, people just want to use something that works so they can spend on their time and resources on building their own apps/services.

time

How many is on the DevRaven team?

I am a solo founder, but will be onboarding a co-founder very soon.

Love the simplicity !

Reply on news.ycombinator.com