FAQ
We have just released the alpha version of the SDE (June 2024). This version is not certified to handle sensitive data.
We expect to release a beta, which will have been certified for sensitive data, as soon as we can.
Then we will release an MVP, in Autumn of 2024, which we expect to be robust enough to handle the majority of our use cases.
In Spring 2025, we will release the first production version, by which time we expect the SDE to be mature.
Development and bugfixes will continue after that for at least another year.
If you see a feature missing that would be useful for you, contact the support team to request it. Even if you know it's in the plan, knowing who wants it, how many people want it, and when they want it by, will help us prioritise tasks.
We are not charging for access to the alpha release of the SDE. We do not expect to start charging until we release the MVP.
You can see from the dashboard that there is a cost shown in the corner of the tile for each workspace you have access to. This cost is taken from the Azure Billing API, so is an accurate representation of the total cost of your workspace to date. If you drill into your workspace, you can see similar figures, showing how much each of your services is costing you.
Bear in mind that some billing items can take time to show up, maybe up to a few hours, or even a day in some cases, so there may be a slight lag in updating numbers.
Also, the cost API doesn't always work, so these numbers may not be shown from time to time. If that happens, just refresh the page.
Not at this point. Cost estimation is something we're aware is needed, and we intent to provide estimates eventually, but there's no concrete plan yet.
Not yet, but that's also a feature we realise people may want.
The bare workspace doesn't cost much, so there's not much point in trying to save money at the workspace level. The workspace resources, however, can cost more. If a resource looks to be expensive, and you know you're not going to use it for some time, you can 'Disable' it from the resource menu. When you need it again, just re-'Enable' it. Resources can be disabled at the service level (e.g., Guacamole), or at the individual resource level.
For virtual machines, however, the process is more complicated. You first need to go to the Actions menu and select Stop, before disabling. If you omit this step, the VM stays running, and continues to incur costs. When you want to use the VM again, enable it, then choose Start from the Actions menu.
Disabling and re-enabling can take some time, up to 10 minutes or more, so be patient.
The best way to use Python is via conda, which is installed on all our custom images. You should create one or more conda environments for your work, preferably on the shared storage disk so that it can be shared between machines. That way, you won't need to re-create the environment every time you boot up a new VM. E.g., on Linux:
mkdir -p /shared-storage/conda
conda create -p /shared-storage/conda/my-env
conda activate /shared-storage/conda/my-env
conda install numpy
...
There's a problem with the Nexus repository mirror at the moment which means it sometimes hangs. When that happens, you won't be able to pull in new software packages, and any new Linux VMs you create will not be able to connect to the shared storage.
Windows VMs are not affected to the same extent, though once the Nexus server goes down, they won't be able to pull new packages either.
Send us an email, we can reboot the Nexus server. That usually fixes things.
First, please read this documentation and check to see if you find your answer. If not, check out our support channel at https://github.com/Barts-Life-Science/Support.