Storage usage on the Galaxy Cloud console

1

Hello,
I’m new to Galaxy, and especially new to using cloud resources. I set up a cloud instance of Galaxy, but am 100% completely lost. There are a few thing things that don’t really seem to be adding up. I’m very sorry for, what are probably trivial questions, but I would appreciate some help,

So in the cloudLaunch “My Appliances” page, it says I’m running a genomics virtual lab on the AWS cloud. When I click on the access link, it takes me to the GVL dashboard where it says that Galaxy isn’t running, but Cloudman is. So I click on the Cloudman access link, and it tells me that I’m running out of disk space. Now, I managed to get Galaxy working a few weeks ago (not sure how), and managed to download a fastq file that was probably about 6 GB. But the Cloudman console was telling me that I was running out of space. I deleted the file, and then the history pane on Galaxy showed that my history was taking up 12 GB of storage, but hte number on the top right of the console said I was using about 6.8 GB. Then, under cluster status in the cloudman console, it says I’ve used 54 GB / 54 GB, but in the admin console, under file systems, it’s telling me that the different fiel systems are taking up almost 150 GB of storage. So I’m having trouble making sense of all this information. All I did was download a 6 GB file, which I then deleted. So why am I getting all this mixed information?

I’m sorry for the mess of information, but I would greatly appreciate it if someone could point towards something that could help me understand what’s going on. How do I make sense of the memory usage?

Also, I tried restarting Galaxy from the cloudman admin console, but it seems to be taking a long time. Is this normal?

1 Like
2

Sorry to hear about the storage troubles. Hopefully together we can figure out what exactly is going on. So, by default, Galaxy does not really delete files from disk when a file is deleted from a history. It just marks a dataset as deleted in case you change your mind about wanting to un-delete later. To actually remove a dataset from disk, it is necessary to purge it, which is available under the history settings cog by clicking Purge Deleted Datasets.
A more system-level approach would be to ssh to your instance and run the cleanup scripts, as described here: https://galaxyproject.org/admin/config/performance/purge-histories-and-datasets/

Overall, the reason the CloudMan console is showing more disk usage than what you have uploaded to Galaxy is because of Galaxy and other supporting/accompanying software that is running on the same disk.

Finally, if all you have done is uploaded a single file to your instance and realized the disk is too small, I would recommend you just delete that particular instance and launch a brand new one with a larger disk. Things will just go a lot quicker and smoother than fighting with the storage and/or waiting for a disk resize (which can take many, many hours to complete in the case of Amazon).

1 Like
3

Thank you. I’ve deleted the instance and will try again.

1 Like