Optimizing Downtime Management for Integrated CMS and Webstudio Workflows
Optimizing Downtime Management for Integrated CMS and Webstudio Workflows
At a glance
The community member is experiencing downtime issues when updating their Baserow instance, which is linked to their Webstudio website. The comments suggest that this is because the community member is mixing a cloud-based Webstudio with a self-hosted Baserow solution, which can lead to downtime when the CMS is updated.
The community members discuss several options to address this issue, including:
- Using a cloud-based solution that handles downtime and upgrades automatically
- Implementing a graceful upgrade process for the self-hosted CMS, such as spinning up a new instance and switching over to it
- Minimizing downtime to a few minutes
- Lowering expectations or using a cloud-based solution
There is no explicitly marked answer, but the community members provide suggestions and guidance on how to handle the downtime issues.
Hey guys! How'd you deal with downtime of your CMS that's linked to Webstudio? Sometimes I like to update my Baserow instance, but then beside Baserow the website on Webstudio will be down as well for some time. What would be best practices for this? Especially when sites start to get in production. Separating servers? Curious how you guys deal with this.
Well or find a way to do graceful upgrades of your cms instance, e.g. starting a new instance on a separate machine, booting it up, then switching over webstudio to the new instance, then killing the old cms instance.
The later is relatively doable if you don't have writes coming from your frontend. If frontend has ability to post things to your database, then after failover you willl need to get the data that was potentially submitted as you were switching over.
I see, I will need to learn how to do a graceful upgrade.
Yes it's logical. Just curious what options I have in this. Whether I learn graceful upgrades and/or use something like "Always Online" from cloudflare.
Cloud is not really an option, at least when it comes down to Baserow and it's API limitations for their cloud.
I understand that building clouds that go never offline is a complex task, that's where certain network engineers come in I guess!
