Hey folks, we've been running Prefect cloud now fo...
# prefect-community
Hey folks, we've been running Prefect cloud now for a while and find it really good. I'm now considering looking in to self hosting the Server and UI on a K8 cluster we have as opposed to paying for cloud. Im a little confused as to what you get for the ongoing monthly cost when you could host for free. I'd be keen to get others thoughts on the community here as to their experiences. I am aware that we would obviously not have to pay for hosting or updates etc, but given the pricing model this feels like a negligible cost in relation to the monthly fees for cloud. I feel like I could be missing something that could be obvious to the community.
I’m sure you’ll get some other responses, but I can give you my brief perspective: • Server is complicated. There are several services interacting and issues in communication across pods can be hard to debug. • Cloud is more performant. We’ve optimized routes and services to handle higher loads. • Server requires an engineer to upgrade / maintain / etc. I’ve seen people move to Cloud because this cost exceeds the cost of Cloud. • Database performance in Server may degrade over time without maintenance. • Cloud has more features, e.g. SLAs, automations • Cloud has authentication and access controls — Server requires a VPN for security.
☝️ 1
💯 1
Cool, thanks @Zanie. Its just unclear what the SLAs are around cloud in terms of uptime, usually for PaaS you see 9x.x% estimations, just haven't seen anything like that mentioned. Given a quite steep cost when moving to standard, it makes me consider whether hosing this ourselves would be a more financially efficient option, as opposed to just simply staying on the almost 'free' tier of cloud we are on now without support. Absolutely not adverse to paying money at all at this organisation, I'd just like more detailed metrics about what we get for our buck e.g. SLAs, Support over and above the basic Slack, Discourse etc. Primarily it would be good to know: • SLA in terms of uptime • SLA in terms of issue resolution • Support hours (geographically, we are UK based) This is a large global company, so obviously they require these sort of answers.
Ah yeah I was referring to our “SLA” automations feature such as notifying you if your agents go down or if your flow does not finish or start in the expected time. You can contact
for more details on uptime/issue resolution/support.