0
I Use This!
Activity Not Available

News

Analyzed almost 1 year ago. based on code collected about 1 year ago.
Posted about 2 months ago by Kyle Tee
👍 @kylefmohr this makes sense & I like the new "Original Allocation" tooltip way of capturing that information. 📝 (I'm also just noting that the meter widget behaviour stays the same, so when we are at "0 SUs remaining", the meter is full. For me, that's good, it feels intuitive.)
Posted about 2 months ago by Kyle Tee
Hi! I've found myself confused by the way allocation use is displayed. Specifically, I keep wondering whether I have, for example, used 43,814 SUs so far, or if I instead have 43,814 SUs remaining. It might have to do partially with the color ... [More] scheme, but I think it's good and makes sense that it uses Indiana University's colors. Instead what this MR proposes is to change the wording, so in my screenshot above, it would instead read 6,186 SUs remaining, and display the amount of the original allocation in the tooltip, under the start and end dates. Here's a mockup: I'm open to any suggestions/changes, and of course I will totally understand if you think it should stay the way it is now. Thanks, and have a good weekend! [Less]
Posted about 2 months ago by Kyle Tee
☑️ Moved.
Posted about 2 months ago by Kyle Tee
I've updated the MR with: 2801d09c feat: choose port range type any, single, min-max in rule form f0496317 refactor: security group rule display direction “incoming” & “outgoing” 169f35fc feat: while editing a rule, de-emphasize the create ... [More] ruleset button 311c56ec feat: start editing a new security group rule when it is added I think some UI which would resolve this is to replace the row with the rule form in the table. is still a WIP. [Less]
Posted about 2 months ago by Frank Tackitt
Overview Complementary work for !984. Depending on their remote types, Security Group Rules can take remote IP prefixes as a parameter. The remote IP prefix is a subnet prefix in CIDR notation. This adds helpers to: Check whether a CIDR is valid. ... [More] Expand IPv6 addresses from their compact form to full form e.g. 2001:db8::1 -> 2001:0db8:0000:0000:0000:0000:0000:0001. How to Test Review the test cases in Tests.elm. Run npm run test. [Less]
Posted about 2 months ago by Julian Pistorius
Posted about 2 months ago by Chris Martin
We already support Nova's "locked" semantic for instances. Does it make sense to look at this holistically for all the resource types that Exosphere handles? Manila shares support resource locks (with the ability to specify an explicit reason!) ... [More] Volumes don't appear to support a locked or protected status. We could achieve this synthetically using volume metadata (locked: true), it will only be meaningful to Exosphere, but probably better than nothing. Floating IPs and SSH public keys: eh. They don't store much data and are usually easy to re-create. [Less]
Posted about 2 months ago by Chris Martin
What would success / a fix look like? Easy for a workflow packager to create executable versions. It is possible to automatically create an up-to-date, executable version of a workflow from a declarative description. It is possible to ensure that ... [More] a workflow is still functioning as intended. Ideally old snapshots are archived in such a way that it is possible to inspect the differences between versions, and possibly even run an older version (e.g. in case something breaks during an update of a workflow). It's unclear what a solution to this issue would actually look like in Exosphere. The path we have explored so far is REES a.k.a. creating an instance with a 'data science workbench'. We've also talked about letting people create instances that run custom Terraform/Ansible code on first boot (i.e. #875). Exosphere doesn't try to help people build, version, or maintain declarative workflow descriptions, nor is it clear what would make sense to do, given Exosphere's role as a client for OpenStack-based clouds. That feels to me like a job for other tools. This issue might be poorly-scoped given what Exosphere is and isn't. I think there may be a lot of value in futher refinements to the data science workbench feature, but that's mostly on the workflow/workbench consumption side, not the authoring or maintenance side. [Less]
Posted about 2 months ago by Chris Martin
In my experience, this feature isn't especially valuable to implement. At risk of jinxing it, we very seldom have someone with a client-side or browser issue that we need to troubleshoot. But my visibility is quite limited. @julianpistorius and his ... [More] RCS colleagues answer all the Jetstream2 support tickets. If I'm wrong and there are frequent user issues with Exosphere itself that we cannot easily identify or reproduce, please correct me. [Less]
Posted about 2 months ago by Chris Martin
This seems like an issue to resolve with tutorials or documentation outside the Exosphere project, so I'm closing it. Someone re-open if you disagree.