[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: [Pulp-list] Creating RHEL 5 compatible yum repositories



Hi Tim,
 
The way I fixed the checksum for my installation of 2.1.3 was to manually edit the python file - on my system:
/usr/lib/python2.6/site-packages/pulp_rpm/yum_plugin/metadata.py
 
I changed all the instances of "sha256" to "sha".  This just forces all repos to use SHA for all checksums - otherwise I was getting a mix of SHA and SHA-256.  I think part of the issue was actually with the modifyrepo command being used, which doesn't seem to have an option for specifying the checksum type.
Christina
On Mon, Dec 16, 2013 at 1:23 PM, Baird, Josh <jbaird follett com> wrote:
Hi,

This was fixed in 2.3.0 [1].  We are successfully syncing RHEL5 repositories with 2.3+.

[1] - https://bugzilla.redhat.com/show_bug.cgi?id=1029057

Thanks,

Josh

-----Original Message-----
From: pulp-list-bounces redhat com [mailto:pulp-list-bounces redhat com] On Behalf Of Tim Bielawa
Sent: Monday, December 16, 2013 1:18 PM
To: pulp-list redhat com
Subject: [Pulp-list] Creating RHEL 5 compatible yum repositories

Prelude: This is in response to BZ1042208 [0], "rpm metadata ignores --checksum-type from repository"

Has anyone else experienced issues/victories using pulp 2.1+ to create YUM repositories which are compatible with RHEL 5?

Specifically, I am referring to pulp setting/not setting the checksum type of a newly created repository correctly?

I have a large number of RHEL 5 boxen to maintain and when we migrated to Pulp 2.1 from 1.x we began seeing an issue where:

> pulp-admin ... rpm repo create --checksum-type sha ...

does not honor the '--checksum-type' option [1], and instead creates all new repositories of type SHA256. This causes a backwards compatibility issue with our RHEL 5 boxes who only support the SHA checksum type.

I am aware of the python-hashlib package, and have verified that installing it successfully allows RHEL 5 boxen to consume repositories of the SHA256 checksum type. However, IMHO this is a non-obvious work around, and not a change I (or other system administrators) can potentially make across my entire RHEL 5 infrastructure without a lot of coordination and approvals.

Thanks for any advice!


[0] https://bugzilla.redhat.com/show_bug.cgi?id=1042208
[1] https://github.com/juicer/juicer/issues/202

--
Tim Bielawa, Senior Release Engineer/Scribe, Inception Why Should I Care What Color the Bikeshed Is?
         http://bikeshed.com/
919.332.6411 Cell | IRC: tbielawa (#inception)
1BA0 4FAB 4C13 FBA0 A036  4958 AD05 E75E 0333 AE37

_______________________________________________
Pulp-list mailing list
Pulp-list redhat com
https://www.redhat.com/mailman/listinfo/pulp-list


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]