Over half way through winter and spring is nearly on-premises. While the temperatures have been icy, come get some hot new insights from our vendor portal. It’s time for another fresh batch of exciting capabilities newly available in our application manager, our Kubernetes installer, and the vendor portal. Check out the recently shipped features and release highlights for February 2023 below.
The instance detail page helps vendor support engineers, product engineers, and anyone else on a vendor team who might be jumping into a customer engagement, quickly gain context on
To read more about this exciting new set of features be sure to read our blog post on Replicated Insight Key Metrics
Previously, pod logs containing very long lines could bloat support bundles to an unreasonable size. The size of each pod log is now limited to 5MB (configurable), in addition to the existing limits of number of lines and age.
If you want to customize the number of bytes are collecting for a specific log file. The example below changes this log file size to collect 10mb.
A vendor’s branding, and the security review process they go through with customers, can be improved by removing references to Replicated from the installation of their application. A vendor’s own domain can be used to replace references to registry.replicated.com when images are hosted in the Replicated registry, and proxy.replicated.com when the app manager proxies images from an external private registry. See this this demo for more about how to use this feature yourself.
Troubleshoot now as the availability to test UDP port status. If you have an application using UDP be sure to leverage this new functionality to enhance your supportability. See the details here.
The sysctl collector now includes virtual memory parameters. Files which previously generated ‘noise’ in results, e.g. the ‘previous’ log files from pods, can now be excluded from analysis so the results are much cleaner.
Both the Postgres and Redis collectors now support connections with TLS. Vendors wanting to use preflights and/or collectors for Redis and Postgres databases that require TLS can now use the collector.
Previously pod logs were stored in various directories within a support bundle depending on the path defined in the logs collector spec. This made it hard to discover logs, and for sbctl to show them as users expected. With this change logs are now stored in the same root directory to work with sbctl, which allows sbctl to return collected pod logs from more than just the cluster-resources collector, better meeting user expectations of how sbctl should work. For backwards compatibility, symlinks to logs are created to allow people to find logs where they used to, without needing to store those logs twice.
Files which previously generated ‘noise’ in results, e.g. the ‘previous’ log files from pods, can now be excluded from analysis so the results are much cleaner.
If you’ve had noisy log files you want to exclude from your analyzer, you can use this to make sure the analyzer providing value.
This migration allows a seamless transition from weave to Flannel, our new CNI of choice, for all Kubernetes installer clusters.
Update your installer specifications today to receive the latest supported networking technologies that Flannel brings to your application.
This update brings kurl up to with the very latest rook version. This update is a rolling update that takes you through each minor release until you reach the latest version. Before your customers upgrade, if they were using file backed devices in rook (pre Rook v1.3), they will need to have a unmounted block device available, as newer versions of rook do not have host file device support. If your app doesn’t need replication, hold tight for a coming details on a migration to OpenEBS that won’t require the block devices.
Vendors have inquired about how customers can generate a merged support bundle, which is a support bundle generated from multiple sources (e.g., a URL, file, ConfigMap/Secret). To reduce support tickets and to help vendors self-serve their customers, we added some conceptual and procedural information in the Enterprise documentation.
The vendor documentation now contains some examples about how vendors can use a modular design approach for support bundle specifications, including using online specifications and making them discoverable. This supports our goal of helping vendors to onboard successfully with Replicated.
Follow these instructions to make your support bundle generation more simple and efficient!
Come see our new table of contents designed to reduce the number of clicks necessary to find the information you need.
We re-wrote the documentation about how to update the version of Kubernetes, the app manager (KOTS), and any kURL add-ons in embedded clusters to provide more clear, detailed instructions and conceptual information. If you are using embedded clusters be sure to read the latest documentation on keeping clusters up to date.
That’s it for the February release highlights! Want to learn more about these new features and what Replicated does to help vendors and customers install and manage modern apps on-prem? We would love to show you -- click here to schedule a demo.