I am trying to run galaxy through the genomics virtual lab on AWS, install some tools, and share it as a share string so that next time I spin up a galaxy instance the genomics virtual lab I can use the share string so the galaxy that spins up already has these tools installed. My question is, is that what a share string should be used for and how it should work? If so, the share button that the support doc says should be next to my cluster name is not there, any advice on why that might be?
Yes, that’s how the functionality is envisioned. Unless you want to share the cluster with someone else, you don’t need the share string for the installed tools (and data) to persist between shutdowns. And that brings me to the reason why the share cluster icon is not showing up: for either of these two options to work, it’s necessary to specify the Persistent Volume Storage option under the Advanced GVL options toggle on the second page of the launch wizard. Once selected, CloudMan will create a persistent disk to which all the data will be placed, which will in turn enable cluster sharing.
Thank you for the reply,
Just to be clear, you are saying if I add persistent volume storage, I would then not need to share the cluster for the installed tools to persist between shutdowns? I thought every time you run a cluster, it is a totally different instance of galaxy?
If you launch a cluster with the persistent volume storage and just shut it down (instead of deleting it), you can choose to relaunch an existing cluster from a drop down menu in CloudLaunch the next time you’d like to use the cluster. The drop down menu shows up on the second page of the GVL launch wizard, after toggling the Advanced GVL options and then clicking the Fetch saved clusters button.
In that case, the same cluster will be relaunched.
Great, that seems to work! Thanks for the info. By not deleting it, you mean not hitting “delete permanently and archive” on the deployment right? So I can can still archive the original deployment so that my active appliances won’t get cluttered with every time I run an instance?
Ahh, for the persistent clusters, it’s necessary to use the CloudMan console to shut down a cluster instead of using the CloudLaunch portal. There, you’ll see a shut down button and a popup after you click the button asking if you’d like to delete or just shut down.
Once you’ve shut down the cluster via CloudMan, you can archive it in CloudLaunch to clean up that page. A new one will show up after you relaunch a new instance.
Got it, so basically
- Start an instance with persistent volume storage
- Add some tools or configure that Galaxy how I want it
- Shut it down but not delete it on the Cloudman console
- Archive it on Cloudlaunch
- Next time I want to use that customized Galaxy instance, find it in “fetch saved clusters”.
Thanks for your help