I think there's still a big market gap around an easy to use, high-level PaaS like Heroku but not sure this quite fills it.
Besides getting jammed into Salesforce, it seems like adding data warehousing, job queueing (besides basic Redis), and better metric support onto Heroku are common growth pains for startups.
Cursory glance: pricing seems reasonable. That's probably going to be a tiny fraction of whatever you're going to pay the engineer(s) who use it, anyway
Northflank offers a free tier including Redis (0.1vCPU, 256MB, 4GB disk) as well as PostgreSQL, MongoDB, MinIO and more. Heroku at scale makes little financial sense for many businesses - our pricing is fair so you can scale as and when your infrastructure demands f.e 7GB on Heroku Redis costs $750 vs $72 on Northflank. (disclaimer northflank co-founder)
I'm tiny fish when it comes to hosting (I spend about $50/mo) and I'm slightly between Heroku and DO's pricing.
Northflank appears to be slightly cheaper than Heroku across everything, but its not really worth the squeeze to save $5-10/mo.
If I ever grow out of Heroku, DO's 1GB RAM for $5/mo is way cheaper. In theory, I could host my own redis for $5/mo at 1GB too. I wouldn't trust myself to self-managed a DB though.
agreed - co-founder http://northflank.com here. Heroku has slowed innovating where expectations of infra have changed. You want more power and flexibility from your PaaS - finer-grained access to metrics & logs is one step - we have added LogQL support. On more optionality on queues - we've started with Redis and managed RabbitMQ is on the way.
Whether it's API support, collaborating in real-time, or scheduling cron jobs we're working hard to make PaaS much more powerful.
Other than more queues & metrics what would you like to see in a high-level PaaS?
In particular, I really like how the build services work in NF. Dockerfile as the central focus and very easy configuration even for relatively complicated build setups.
Besides getting jammed into Salesforce, it seems like adding data warehousing, job queueing (besides basic Redis), and better metric support onto Heroku are common growth pains for startups.
Cursory glance: pricing seems reasonable. That's probably going to be a tiny fraction of whatever you're going to pay the engineer(s) who use it, anyway