Each Galaxy server is independent. How much disk space to allocate for end users is determined by the administrators of that server.
If you are working at Galaxy Main (http://usegalaxy.org), each registered account is granted 250 GB of quota space (disk usage for data storage). Review full details here (account and job quotas) and here (account terms).
Should you have questions after reviewing the options below, this is how to get help:
- Review prior Q&A or ask a novel question at the Galaxy Help forum https://help.galaxyproject.org/.
- Chat with us at Gitter https://gitter.im/galaxyproject/Lobby/.
- Send an email to firstname.lastname@example.org from your registered account email address.
- All options https://galaxyproject.org/support/#help-resources.
Working at a different Public Galaxy server? If you question is general, try the Galaxy Help forum. If server specific or account related, that server's home page or directory page usually contains contact and other information concerning account terms, questions, issues, and quota policies: Galaxy Platform Directory. If you cannot find the contact information, or are not sure about what the problem is exactly, ask for advice at the Galaxy Help forum or at Gitter. The more context you provide (server URL, problem details), the quicker our community will be able to assist you.
Log out of Galaxy, then back in again. This refreshes the disk usage calculation dislayed in the Masthead usage (summary) and under User > Preferences (exact).
The options are included under the User menu in the top masthead.
If the account usage is showing that quota is exceeded -over 250 GB:
- All account Histories can be reviewed under User > Histories.
- Only active histories are displayed by default.
- Click on Advanced Search, then choose status = all to review all active, deleted, and purged histories.
- Locate deleted but not permemently deleted histories and purge them.
- Download any datasets or entire histories you want to keep first. The history will need to be in an active state. Histories that are deleted can be undeleted, but once a history is purged (aka permenently deleted) it cannot be recovered, even by an administrator.
- Histories must be unshared before they can be deleted or purged. Unshare first as needed.
- Histories in a deleted state are still part of your quota usage.
- Histories must be purged (permenently deleted) to not count toward quota.
Find all Datasets or just review those in the active history and purge (aka permenently delete) those not needed
- All account Datasets can be reviewed under User > Datasets.
- Within a History, deleted/permenently deleted Datasets can be reviewed by toggling the deleted link at the top of the History panel, found immediately under the History name.
- Both active (shown by default) and hidden (the other toggle link, next to the deleted link) datasets can be reviewed the same way.
- Click on the far right "X" to delete a dataset.
- To permenently delete: use the link from within the dataset, or use the Operations on Multiple Datasets functions, or use the Purge Deleted Datasets option in the History menu.
- Datasets in a deleted state are still part of your quota usage.
- Datasets must be purged (permenently deleted) to not count toward quota.
- All account Histories owned by others, but shared with you, can be reviewed under User > Histories shared with me.
- Review the list. There may be none!
- Only a small fraction of the data content in "Histories shared with me" are part of your own quota usage, and unsharing will not significantly reduce quota usage for most end-users. However, if there are hundreds (or more!), or many large histories shared with you, clearing these up periodically (unsharing) could be worth it.
- The other person does not need to unshare a history for you. Unshare histories yourself on this page using the pull-down menu per-history.
- View is a review feature. The data cannot be worked with but many details, including tool and dataset metadata/parameters are included.
- Copy those you want to work with. This will increase your quota usage since that history copy is now one of your histories, but that will also allow you manipulate the datasets or the history itself, independently from the original owner. All History/Dataset functions are available if the other person granted full access to the datasets to you. Dataset sharing is a distinct option on the Share or Publish form. Only the original History owner can set the level of sharing (History only, or also Datasets). If you need Dataset access and it wasn't granted, you might need to ask the other person to reshare with the box for "Also make all objects within the History accessible." checked. The other person can toggle sharing to "off" for review/changes, then toggled back to "on" after.
- Unshare any on the list not needed anymore. After a history is copied, you will still have your own version of the history, even if later unshared or the other person who shared it with you makes changes to their own version of the history later on. Meaning, each account's version of a History, and the Datasets in it, are distinct (unless the Datasets were not shared -- then you will still only be able to "view" Datasets, but not work with or download them).
- Dataset and History privacy options, including sharing, can be set under User > Preferences. The automatic default is that none of your account data is shared. You must actively choose to share your work -- both the privacy level AND how -- for any Galaxy Object (Dataset, History, Workflow, Visualization).
- A small fraction of space from "Histories shared with me" in a shared state are still part of your quota usage.
- Shared Histories must be unshared to not count toward quota usage.
- If you share a History with someone else, that does not increase OR decrease your own quota usage.
- There are several options to preserve data/work while also reducing your quota usage. Some examples are below.
- Download Datasets as individual files or entire Histories as an archive. Then purge them from the public server.
- Copy Datasets or Histories to another Galaxy server, including your own Galaxy. Then purge.
- Copy your most important Datasets into a new/other History (inputs, results), then purge the original full History. This can be a quicker alternative if the orignal History is very large or complex to review/purge dataset-by-dataset.
- Extract a Workflow from the History, then purge it. Workflows do not consume quota space as they do not contain any datasets, only tools/parameters. IF you first saved back the original inputs and final result datasets, the analysis can be rerun to recreate the intermediate datasets and to compare the new result with your original. Be aware that minor differences are expected over time, as tools and Galaxy itself are updated.
- Back-up your work. It is a best practice to download an archive of your FULL original Histories periodically, even those still in use, as a backup.
- In short, keep active work on the server and archive/download older work + backups of current work. You can always reupload prior downloaded Datasets or Histories later. Or, rerun an analysis with a Workfow and the original inputs.
- Resources Much discussion about all of the above options can be found at the Galaxy Help forum https://help.galaxyproject.org/ (in context with difference use cases). The Support FAQs and Admin Docs are more places to find help. If any option is unclear or you run into problems, ask for more help! https://galaxyproject.org/support/#help-resources
Have a very large history that is difficult to mange? Review the options below for some tips.
- Navigate to a history (User > Histories) and use the history menu function "Purge Deleted Datasets". That will make sure that any datasets that were only deleted become permanently deleted (purged). Note that purged data cannot be recovered, even by an administer.
- Navigate to a history (User > Histories) and use the history menu function Copy Datasets. Only active datasets will appear in the middle panel view listing of datasets. Select the datasets that you wish to retain and copy those into a new history. Once completed, go back and permanently delete (purge) the entire original history under User > Histories. You may need to log out then back into Galaxy again to reset the quota usage.
- If you copy a dataset collection to a new or different history, the hidden datasets associated with that collection are also copied.
- Copies of datasets within your own account do not individually count against quota usage -- these are "clones" unless a change is made (after a change, copies do become distinct datasets and count toward quota usage individually). If you change a copied dataset, consider updating/removing the original tag(s), or add more tag(s).
- Keep future histories smaller by breaking up analysis into different histories. You can always keep copies (aka "clones") of important or frequently reused datasets datasets in different histories. Examples include: reference annotation, custom fasta genomes/transcriptomes/exomes, processed results, original source read data (pre or post QA/QC).
- Sometimes it is helpful to add tags to datasets that are important or frequently reused to better keep track of datasets that represent the same original data in different histories. Tag the original dataset or dataset collection (#name-of-tag) and any copies will inherit that same tag when copied.
- Consider creating a dedicated history for important or frequently reused datasets. You can create a copy of that entire history to a new history, give it a meaningful name, then start a new analysis.
- Histories can also have tags added to help you keep analysis/data organized.
- To review all of your histories, go to User > Histories. Note that if you have many histories, that view may take some time to load. Be patient, it will load, and when you reload that view again within a short time period, the view will load/resolve quicker. The listing includes the history name, dataset state summaries, overall size of active + hidden + deleted datasets, associated history tags, sharing status, and the history state. Only active histories are shown by default. Use the Advanced Search function to review all histories with different states (active, deleted, permanently deleted aka purged). Only purged datasets and histories do not count toward quota usage.
- To review all of your datasets, go to User > Datasets. Note that if you have many datasets, then that view may take some time to load. Be patient, it will load, and when you reload that view again within a short time period, the view will load/resolve quicker. The listing includes the dataset name, associated tags, the datatype, and when the data was last updated. The view can be sorted by any of those attributes. You can also search and copy datasets from that view, plus navigate to the history where a dataset is located. Note: Only active datasets are displayed on that view -- hidden, deleted, and purged datasets are not listed and dataset cannot have the state changed. You must manage a dataset state from within the history containing the dataset.
Account quotas are generally fixed at 250 Gb for Galaxy Main. Other Galaxy servers may differ for quota size and policy. For very large projects, consider moving to a local or cloud Galaxy where you can control the resources.
Important: Increased working space (quota size) is unrelated to the computational resouces used to process analysis jobs. If your jobs are failing because they are exceeding resources, there is either an input/usage problem or the work is actually too large to process at the server. Not sure? Check here: My job ended with an error. What can I do?.
- Some work is simply too large for the public resource. How to know? Review the Using Galaxy choices explained here and the Webinar: Use Galaxy on the web, the cloud, and your laptop too.
- If your work is academic research based and additional short-term space is needed at Galaxy Main (https://usegalaxy.org), please send an email to email@example.com from your registered account email address, briefly explaining your project and we may be able to help.
- Also include this information in your email or we will request it, delaying action: 1) your primary academic email address 2) how much space you need 3) how long you need the extra space.
- Double-check that you are actually purging old work correctly. We will check first before granting any extra quota.
- Note: If you have registered or are using or have used more than one account at the Galaxy Main server, this violates our Terms of Usage. Account terms are explained in a simple and specific sentence exactly where you clicked on the activation link in the registration email received when the account was first opened. If that was not clear enough, you must proactively clear up account issues and get down to and keep using just one account before extra quota can be granted. Yes, we will check when you ask, and deny requests that are incomplete. Duplicated accounts created/detected after extra quota is granted will result in the loss of all associated accounts and data, without another reminder.
- Not sure what to do? If you have more than one account (old or new) at UseGalaxy.org, manage the data yourself and delete extra/old accounts under User > Preferences. Review here if terms of usage are not clear and email us to sort things out if you need to: Getting an account at Galaxy Main (http://usegalaxy.org)
- Other Public Galaxy server's may have different account terms or quota increase request proccesses. Check the server's home page or the server's directory page here to locate contact and other information concerning account terms, questions, issues, and quota policies: Galaxy Platform Directory