You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Current deployments of conda-store require manual intervention to recover disk space when many users create many environment builds, because conda-store keeps old builds indefinitely. This issue is to discuss proposals for implementing a format by which a old build can be archived when not in use, and recreated if a user requests for it to be activated again.
Value and/or benefit
Currently, keeping every build keeps every package in each build. In cases where many old build exists, this accounts for the majority of the disk space used by conda-store, when only the most recent build or two for each environment is actually being used. Archiving old builds would recover most of the disk space while still allowing users to revert if need be, greatly reducing the maintenance burden for existing conda-store deployments.
Related issues
See also #791, which is a proposal to update the default conda-store storage limits. While that approach could alleviate some of the pressure on existing deployments, a long term fix will involve more robust archiving/recreation capabilities.
The text was updated successfully, but these errors were encountered:
Feature description
Current deployments of
conda-store
require manual intervention to recover disk space when many users create many environment builds, becauseconda-store
keeps old builds indefinitely. This issue is to discuss proposals for implementing a format by which a old build can be archived when not in use, and recreated if a user requests for it to be activated again.Value and/or benefit
Currently, keeping every build keeps every package in each build. In cases where many old build exists, this accounts for the majority of the disk space used by
conda-store
, when only the most recent build or two for each environment is actually being used. Archiving old builds would recover most of the disk space while still allowing users to revert if need be, greatly reducing the maintenance burden for existingconda-store
deployments.Related issues
See also #791, which is a proposal to update the default
conda-store
storage limits. While that approach could alleviate some of the pressure on existing deployments, a long term fix will involve more robust archiving/recreation capabilities.The text was updated successfully, but these errors were encountered: