git failures

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
12 messages Options
Reply | Threaded
Open this post in threaded view
|

git failures

Andrew Dunstan-8

Recently a number of git failures have been seen like this:


    *** Please tell me who you are.

    Run

       git config --global user.email "[hidden email]"
       git config --global user.name "Your Name"

    to set your account's default identity.
    Omit --global to set the identity only in this repository.

    fatal: empty ident  <[hidden email]
    > not allowed

Notably on Debian/Ubuntu machines. I assume they have had some sort of
recent git update.

Why it should care if you're not committing anything is beyond me. It
seems quite pointless. However, It seems like the account you're running
the buildfarm animal as needs a name and email address set to avoid this
issue.

This applies particularly to members spurfowl, ibex and termite. Please
fix them urgently so the failures email list doesn't get spammed.

cheers

andrew

--

Andrew Dunstan                https://www.2ndQuadrant.com
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services



--
Sent via buildfarm-members mailing list ([hidden email])
To make changes to your subscription:
http://www.postgresql.org/mailpref/buildfarm-members

Reply | Threaded
Open this post in threaded view
|

Re: git failures

Tom Lane-2
Andrew Dunstan <[hidden email]> writes:
> Recently a number of git failures have been seen like this:
>     *** Please tell me who you are.
> Notably on Debian/Ubuntu machines. I assume they have had some sort of
> recent git update.

No, I do not think it's the fault of the buildfarm machines.  The problem
is that Magnus got rid of a bad commit into the 9.2 branch via "git reset"
not "git revert", cf

https://www.postgresql.org/message-id/CABUevEzrD-BYtQOQ22navkoWquOC5TcqPcBS_iQ2cYqaK9_39g%40mail.gmail.com

and these three machines seem to have been quick enough to see the
intermediate state and get confused by it.  I'm not very sure why "git
pull" isn't enough to recover, but it evidently isn't.

                        regards, tom lane


--
Sent via buildfarm-members mailing list ([hidden email])
To make changes to your subscription:
http://www.postgresql.org/mailpref/buildfarm-members

Reply | Threaded
Open this post in threaded view
|

Re: git failures

Stephen Frost
Andrew, Tom,

Out of the office currently but will address as soon as I'm able to.

Thanks!

Stephen

On Tue, Jan 3, 2017 at 17:21 Tom Lane <[hidden email]> wrote:
Andrew Dunstan <[hidden email]> writes:

> Recently a number of git failures have been seen like this:

>     *** Please tell me who you are.

> Notably on Debian/Ubuntu machines. I assume they have had some sort of

> recent git update.



No, I do not think it's the fault of the buildfarm machines.  The problem

is that Magnus got rid of a bad commit into the 9.2 branch via "git reset"

not "git revert", cf



https://www.postgresql.org/message-id/CABUevEzrD-BYtQOQ22navkoWquOC5TcqPcBS_iQ2cYqaK9_39g%40mail.gmail.com



and these three machines seem to have been quick enough to see the

intermediate state and get confused by it.  I'm not very sure why "git

pull" isn't enough to recover, but it evidently isn't.



                        regards, tom lane



Reply | Threaded
Open this post in threaded view
|

Re: git failures

Andrew Dunstan-8
In reply to this post by Tom Lane-2


On 01/03/2017 05:21 PM, Tom Lane wrote:

> Andrew Dunstan <[hidden email]> writes:
>> Recently a number of git failures have been seen like this:
>>      *** Please tell me who you are.
>> Notably on Debian/Ubuntu machines. I assume they have had some sort of
>> recent git update.
> No, I do not think it's the fault of the buildfarm machines.  The problem
> is that Magnus got rid of a bad commit into the 9.2 branch via "git reset"
> not "git revert", cf
>
> https://www.postgresql.org/message-id/CABUevEzrD-BYtQOQ22navkoWquOC5TcqPcBS_iQ2cYqaK9_39g%40mail.gmail.com
>
> and these three machines seem to have been quick enough to see the
> intermediate state and get confused by it.  I'm not very sure why "git
> pull" isn't enough to recover, but it evidently isn't.


Yeah, axolotl just failed :-(

cheers

andrew

--
Andrew Dunstan                https://www.2ndQuadrant.com
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services



--
Sent via buildfarm-members mailing list ([hidden email])
To make changes to your subscription:
http://www.postgresql.org/mailpref/buildfarm-members

Reply | Threaded
Open this post in threaded view
|

Re: git failures

Stephen Frost
All,

I'm hopeful that I fixed it....

Thanks!

Stephen

On Tue, Jan 3, 2017 at 19:09 Andrew Dunstan <[hidden email]> wrote:




On 01/03/2017 05:21 PM, Tom Lane wrote:

> Andrew Dunstan <[hidden email]> writes:

>> Recently a number of git failures have been seen like this:

>>      *** Please tell me who you are.

>> Notably on Debian/Ubuntu machines. I assume they have had some sort of

>> recent git update.

> No, I do not think it's the fault of the buildfarm machines.  The problem

> is that Magnus got rid of a bad commit into the 9.2 branch via "git reset"

> not "git revert", cf

>

> https://www.postgresql.org/message-id/CABUevEzrD-BYtQOQ22navkoWquOC5TcqPcBS_iQ2cYqaK9_39g%40mail.gmail.com

>

> and these three machines seem to have been quick enough to see the

> intermediate state and get confused by it.  I'm not very sure why "git

> pull" isn't enough to recover, but it evidently isn't.





Yeah, axolotl just failed :-(



cheers



andrew



--

Andrew Dunstan                https://www.2ndQuadrant.com

PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services



Reply | Threaded
Open this post in threaded view
|

Re: git failures

Tom Lane-2
Stephen Frost <[hidden email]> writes:
> I'm hopeful that I fixed it....

For the benefit of the other people with this problem ... what did you do?

                        regards, tom lane


--
Sent via buildfarm-members mailing list ([hidden email])
To make changes to your subscription:
http://www.postgresql.org/mailpref/buildfarm-members

Reply | Threaded
Open this post in threaded view
|

Re: git failures

Stephen Frost

On Tue, Jan 3, 2017 at 20:25 Tom Lane <[hidden email]> wrote:
Stephen Frost <[hidden email]> writes:

> I'm hopeful that I fixed it....



For the benefit of the other people with this problem ... what did you do?

I did a git pull in the REL9_2_STABLE/pgsql directory and accepted the merge commit message. 

I'm not convinced that was really the "right" or "correct" thing to do, but it seemed to allow subsequent git pull's without erroring or asking for any user interaction.

Will look into it more tomorrow. 

Thanks!

Stephen
Reply | Threaded
Open this post in threaded view
|

Re: git failures

Andrew Dunstan-8


On 01/03/2017 08:32 PM, Stephen Frost wrote:

>
> On Tue, Jan 3, 2017 at 20:25 Tom Lane <[hidden email]
> <mailto:[hidden email]>> wrote:
>
>     Stephen Frost <[hidden email] <mailto:[hidden email]>> writes:
>
>     > I'm hopeful that I fixed it....
>
>
>
>     For the benefit of the other people with this problem ... what did
>     you do?
>
>
> I did a git pull in the REL9_2_STABLE/pgsql directory and accepted the
> merge commit message.
>
> I'm not convinced that was really the "right" or "correct" thing to
> do, but it seemed to allow subsequent git pull's without erroring or
> asking for any user interaction.
>
> Will look into it more tomorrow.
>


That didn't work for me. I removed that directory altogether, and I
expect the next buildfarm run to rebuild it cleanly.

cheers

andrew


--

Andrew Dunstan                https://www.2ndQuadrant.com
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services



--
Sent via buildfarm-members mailing list ([hidden email])
To make changes to your subscription:
http://www.postgresql.org/mailpref/buildfarm-members

Reply | Threaded
Open this post in threaded view
|

Re: git failures

Craig Ringer-3
On 4 January 2017 at 10:15, Andrew Dunstan
<[hidden email]> wrote:

>
>
> On 01/03/2017 08:32 PM, Stephen Frost wrote:
>>
>>
>> On Tue, Jan 3, 2017 at 20:25 Tom Lane <[hidden email]
>> <mailto:[hidden email]>> wrote:
>>
>>     Stephen Frost <[hidden email] <mailto:[hidden email]>> writes:
>>
>>     > I'm hopeful that I fixed it....
>>
>>
>>
>>     For the benefit of the other people with this problem ... what did
>>     you do?
>>
>>
>> I did a git pull in the REL9_2_STABLE/pgsql directory and accepted the
>> merge commit message.
>>
>> I'm not convinced that was really the "right" or "correct" thing to do,
>> but it seemed to allow subsequent git pull's without erroring or asking for
>> any user interaction.
>>
>> Will look into it more tomorrow.
>>
>
>
> That didn't work for me. I removed that directory altogether, and I expect
> the next buildfarm run to rebuild it cleanly.

It's definitely not the right thing.

You should remove the clone, or

git reset --hard origin/REL9_2_STABLE

it.



--
 Craig Ringer                   http://www.2ndQuadrant.com/
 PostgreSQL Development, 24x7 Support, Training & Services


--
Sent via buildfarm-members mailing list ([hidden email])
To make changes to your subscription:
http://www.postgresql.org/mailpref/buildfarm-members

Reply | Threaded
Open this post in threaded view
|

Re: git failures

Craig Ringer-3
In reply to this post by Stephen Frost
Termite is fixed.

Reset to 83a25a5209d93615e6d5f3eee4ef649853478498

--
 Craig Ringer                   http://www.2ndQuadrant.com/
 PostgreSQL Development, 24x7 Support, Training & Services


--
Sent via buildfarm-members mailing list ([hidden email])
To make changes to your subscription:
http://www.postgresql.org/mailpref/buildfarm-members

Reply | Threaded
Open this post in threaded view
|

Re: git failures

Stephen Frost
In reply to this post by Craig Ringer-3
* Craig Ringer ([hidden email]) wrote:

> On 4 January 2017 at 10:15, Andrew Dunstan
> <[hidden email]> wrote:
> >
> >
> > On 01/03/2017 08:32 PM, Stephen Frost wrote:
> >>
> >>
> >> On Tue, Jan 3, 2017 at 20:25 Tom Lane <[hidden email]
> >> <mailto:[hidden email]>> wrote:
> >>
> >>     Stephen Frost <[hidden email] <mailto:[hidden email]>> writes:
> >>
> >>     > I'm hopeful that I fixed it....
> >>
> >>
> >>
> >>     For the benefit of the other people with this problem ... what did
> >>     you do?
> >>
> >>
> >> I did a git pull in the REL9_2_STABLE/pgsql directory and accepted the
> >> merge commit message.
> >>
> >> I'm not convinced that was really the "right" or "correct" thing to do,
> >> but it seemed to allow subsequent git pull's without erroring or asking for
> >> any user interaction.
> >>
> >> Will look into it more tomorrow.
> >>
> >
> >
> > That didn't work for me. I removed that directory altogether, and I expect
> > the next buildfarm run to rebuild it cleanly.
>
> It's definitely not the right thing.
>
> You should remove the clone, or
>
> git reset --hard origin/REL9_2_STABLE
>
> it.
Yeah, seems like that was necessary.  I've done that now and it seems
happier.

Thanks!

Stephen

signature.asc (819 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: git failures

Andrew Dunstan-8
In reply to this post by Craig Ringer-3


On 01/03/2017 11:55 PM, Craig Ringer wrote:

>>
>>
>> That didn't work for me. I removed that directory altogether, and I expect
>> the next buildfarm run to rebuild it cleanly.
> It's definitely not the right thing.
>
> You should remove the clone, or
>
> git reset --hard origin/REL9_2_STABLE
>




This seems to be the only reliable fix. Even removing the directory
didn't work where it's a linked workdir.

Thanks, Craig!

cheers

andrew

--
Andrew Dunstan                https://www.2ndQuadrant.com
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services



--
Sent via buildfarm-members mailing list ([hidden email])
To make changes to your subscription:
http://www.postgresql.org/mailpref/buildfarm-members