Each snapshot represents a layer of the docker image, plus there’s one for the container itself. So the number of snapshots resulting from a Rock-on depends on how many layers there are in underlying docker images. There are two ways to reduce number of snapshots. 1) create your images from scratch and have fewer layers or 2) just delete the older snapshots holding data of bottom layers. This is technically safe, and reduces the number of snapshots but does not delete any data. More info in docker docs here. We may want to do this at some point, but at this point we need to do some testing. Please feel free to create a github issue and post your test results.
I managed to merge three pull requests that queued up recently. Each of them improves the UI a little. Thanks to @ganti_priya and @mchakravartula for these contributions.
3.8-13.05 is now available, special thanks to our newest contributor @demount, who ran into a problem, fixed it and opened a pull request! @Flyer added rfc2307 schema support to AD integration with rfc2307 schema support. @ScarabMonkey helped with testing. Thank you all!
Happy sunday everyone! 3.8-13.06 is now available.
@phillxnet added a big feature: HDD power down based on idle time. This is perhaps for the advanced users who like to spin their drives down to reduce noise and power consumption at times. The feature is wonderfully documented on the pull request page. This is a big enough feature that deserves more documentation and tutorials in the form of wiki, blog, youtube videos etc… it’s exciting.
Thanks to @maxhq for helping with code review and also for contributing a pull request to sort Shares by name. This default improves UX when choosing a Share, for example: rock-on configuration, samba exports etc…
@KarstenV Thanks for the report and comments and screen shot. By design and until we get a notification system in place we log the reason for no response to /opt/rockstor/var/log/rockstor.log. I suspect that the samsung’s are not being correctly identified as rotational. I wanted to err on the cautious side and do nothing unless we can confirm a device is rotational. In the meantime, as I know you already know, we can watch this log using the following command.
tail -f /opt/rockstor/var/log/rockstor.log
@Flyer is actually making great progress on a log reading facility within the WebUI but for now we will have to command line it.
Could you open a new forum post that we can link to here as I would like to work out why this is happening with these drives. If it turns out that your rockstor.log has the following in every time you try and set the spindown on the Samsungs then we can take it from there.
“Skipping hdparm settings: device not confirmed as rotational”
Could you also in the new thread (as this may take a few posts back and forth) post the the output of:-
udevadm info --query=property --name devNameOfASamsung
Cheers.
By the way the APM setting is blended in with the spindown setting via the slider as sometimes too higher APM setting can inhibit spindown, so I presented the two together.
@ganti_priya refactored a bunch of frontend js templates, making the code more readable. @Flyer fixed a regression in schedule task execution code. Thank you both!
The last commit of this release is 58ee139. Three issues were closed, each of which fixes and improves the UI a little bit. Thanks @ganti_priya and @Flyer for your contributions!
Almost forgot to reply here, 3.8-13.09(5213dd5) was released last night (pacific time).
@phillxnet fixed a Rock-on metadata update bug, @ganti_priya fixed a replication related UI bug and @Flyer fixed a dashboard bug. Thanks for your contributions. pleasure to merge them!
Happy sunday folks! 3.8-13.10(ca1d9b5) is now available.
A total of three issues were merged. Two of them are UI improvements, something @ganti_priya has been methodically working on. Third is an improvement to AD configuration by @ScarabMonkey. Way to go about fixing issues you use and care about! pleasure to merge them.
I am happy to announce that 3.8-13.11(ec65d1c) is now available.
I’ve merged three pull requests in this update. @Flyer’s Logs Manager is the big one. It’s a really cool feature and I encourage everyone to use it and provide your feedback. Thank you @Flyer for sticking with it and doing an impressive job!
@phillxnet made an important improvement to disk state management. We can now store generic auxiliary information and assign roles to HDDs, paving way for features like backup disks etc…
@ganti_priya fixed a replication related UI bug. Happy to merge three pull requests from three different contributors, making this update special! Thank you!
