Bug 807668

Summary: RFE: After metadata download, Pulp should report when is not enough disc space for rpms or files
Product: [Retired] Pulp Reporter: Lukas Zapletal <lzap>
Component: user-experienceAssignee: amacdona <austin>
Status: CLOSED NOTABUG QA Contact: Preethi Thomas <pthomas>
Severity: low Docs Contact:
Priority: low    
Version: MasterCC: austin, omaciel, rbarlow, skarmark
Target Milestone: ---Keywords: FutureFeature, Triaged
Target Release: ---   
Hardware: Unspecified   
OS: Unspecified   
Whiteboard:
Fixed In Version: Doc Type: Enhancement
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2014-11-19 21:05:12 UTC Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Attachments:
Description Flags
pulp.log
none
/var/log/messages none

Description Lukas Zapletal 2012-03-28 12:52:19 UTC 
We hit out of disc space errors way too often while working in Katello and I can imagine there will be users hitting the similar issues.

The problem is sync can be quite long and we should maybe report this in advance to the user.

My request is to implement disc size check before starting downloading rpms or iso files (or where possible) and have an option to disable this behavior. By default it should be switched on.
Comment 1 Og Maciel 2012-03-28 13:03:00 UTC 
Having just gone through a case where I ran out of disk space, I'm attaching the tracebacks found on pulp.log and messages log in case it may provide useful information.
Comment 2 Og Maciel 2012-03-28 13:03:42 UTC 
Created attachment 573348 [details]
pulp.log

Python traceback from pulp.log
Comment 3 Og Maciel 2012-03-28 13:04:14 UTC 
Created attachment 573349 [details]
/var/log/messages

/var/log/messages
Comment 4 amacdona@redhat.com 2014-11-19 21:05:12 UTC 
I do not think it is possible for Pulp to predict running out of disk space, especially since these items are being stored in a database.

When the disk is full, it is Mongo, not Pulp that crashes. Pulp correctly responds that it cannot connect to the database. 

I am closing for now, please feel free to reopen if I missed something.