Bug 17760 - rpm does unnecessary FTP downloads
Summary: rpm does unnecessary FTP downloads
Keywords:
Status: CLOSED WONTFIX
Alias: None
Product: Red Hat Linux
Classification: Retired
Component: rpm
Version: 7.0
Hardware: i386
OS: Linux
medium
medium
Target Milestone: ---
Assignee: Jeff Johnson
QA Contact:
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2000-09-21 03:42 UTC by roystgnr
Modified: 2008-05-01 15:37 UTC (History)
0 users

Fixed In Version:
Doc Type: Enhancement
Doc Text:
Clone Of:
Environment:
Last Closed: 2000-09-21 03:43:00 UTC
Embargoed:


Attachments (Terms of Use)

Description roystgnr 2000-09-21 03:42:59 UTC
running, for example:

rpm -Fvh ftp://linux.rice.edu/pub/redhat/updates/i386/emacs*,

rpm-4.0-0.66 will attempt to download all 50+ megabytes of RPM files,
*before* it checks the version number on each to discover that I already
have the latest versions of all my installed emacs RPMs.

In fact,

rpm -Fvh ftp://linux.rice.edu/pub/redhat/updates/i386/emacs-leim* will try
downloading that RPM, even though I do not have emacs-leim installed on my
system and so could not possibly want to freshen that package.

Running "rpm -qip", on the other hand, downloads only the package headers
necessary to determine the --info query.  Could this behavior be used for
upgrade and freshen commands as well?  When rpm reaches the end of the
headers on the file it is downloading, could it check package and version
information *then*, and so decide whether to continue or cancel the
download before it wastes bandwidth and disk space on the entire file?

Comment 1 Jeff Johnson 2000-10-03 17:17:13 UTC
On install/upgarde/freshen pathways, rpm always downloads the entire package
before
attempting an install, while query aborts. The rationale for this behavior is
that query
failure, due to network transport problems, is not as critical as package
install/erase failure.

If you don't like the behavior, then supply rpm a list of packages as in
	rpm -Fvh `cat <your_list_here>`


Note You need to log in before you can comment on or make changes to this bug.