Hi Marian,
well, the story of EPEL is long and a complicated one. EPEL is not
facing migration: xrootd (3) => xrootd4 => xrootd (4.1) because they
have refused to make the first step (xrootd (3) => xrootd4). This is the
reason why we have decided to redesign the plugin system in xrootd to
make it possible for xrootd 3 and xrootd 4.1 (and any other major
version) to coexist on the same system. EPEL has just done their
migration xrootd (3) => xrootd (4.1).
If it comes to the exact format of the compat packages, my
understanding was that he have had an agreement with Mattias on what
they will look like, but apparently I was mistaken. This, in principle,
should not be a big deal becasue the compat packages should not be
referenced explicitly in any package. Instead the systems using them
should rely on automatic soname dependency resolution in RPM. That said,
we will likely adjust to follow EPEL, because what they have done is
somewhat more convenient and we benefit from being compatible with them.
Cheers,
Lukasz
On 02/17/2015 06:53 PM, Marian Zvada wrote:
> Lukasz, and others:
>
> OSG aims to provide similar build soon and I was assigned to exercise
> this portion of fun:
>
> https://jira.opensciencegrid.org/browse/SOFTWARE-1788
>
> Make long story short from JIRA above citing that we know EPEL is/was
> facing the same xrootd (3) => xrootd4 => xrootd (4.1) upgrade issues,
> just would like to confirm your opion which direction is most welcome
> that OSG doesn't introduce yet another direction of xrootd 4.1 build
> within their own packaging principles and procedures.
>
> For instance, the cern/xrootd packaging now contains both the xrootd 4
> and xrootd 3 sources and builds both in order to provide the compat
> packages but EPEL makes a separate build for the compat stuff.
>
> As I don't usually do builds I'm happy to learn now more but starting
> with xrootd especially due all these version changes may be challenging
> for me. Just avoid confusion and keep right direction in this matter can
> someone point me to the source of rpms with spec file I may use as an
> example and provide xrootd developers comments/opinions on this above?
>
> Thanks,
> Marian
>
> On 2/16/15, 3:08 AM, Lukasz Janyst wrote:
>> -------- Forwarded Message --------
>> Subject: Re: xrootd 4 & xrootd 3 compat packages and dpm
>> Date: Mon, 16 Feb 2015 09:42:36 +0100
>> From: Mattias Ellert <[log in to unmask]>
>> To: David Smith <[log in to unmask]>
>> CC: Lukasz Janyst <[log in to unmask]>
>>
>> ons 2015-02-11 klockan 16:46 +0000 skrev David Smith:
>>> > On 11 Feb 2015, at 16:34, David Smith <[log in to unmask]> wrote:
>>> >
>>> >
>>> >> On 11 Feb 2015, at 16:16, Mattias Ellert
>>> <[log in to unmask]> wrote:
>>> >>
>>> >> What is your status w.r.t. pushing the updates to stable?
>>> >
>>> > Hi Mattias,
>>> >
>>> > I’ve told the dpm testing sites and warned the rest of the dpm
>>> admins to expect xrootd to move to epel stable on or around the 17th
>>> of this month (I supposed the mirror updates introduce some spread in
>>> when the new xrootd packages appear). We suspended our dpm-xrootd
>>> 3.5.2 package move form testing to stable to coincide with that.
>>>
>>> Just to add:
>>>
>>> Thank you for applying the last delay via the override for the xrootd
>>> package.
>>>
>>> We’ve made, tested and been able to distributed our upgrade process
>>> with forewarning to our sites, so I believe things are well
>>> understood.
>>>
>>> Yours,
>>> David
>>>
>>> --
>>
>> The xrootd update has been submitted to EPEL stable.
>>
>> Mattias
>>
>> ########################################################################
>> Use REPLY-ALL to reply to list
>>
>> To unsubscribe from the XROOTD-DEV list, click the following link:
>> https://listserv.slac.stanford.edu/cgi-bin/wa?SUBED1=XROOTD-DEV&A=1
########################################################################
Use REPLY-ALL to reply to list
To unsubscribe from the XROOTD-DEV list, click the following link:
https://listserv.slac.stanford.edu/cgi-bin/wa?SUBED1=XROOTD-DEV&A=1
|