You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@libcloud.apache.org by Tomaz Muraus <to...@apache.org> on 2016/01/13 09:52:06 UTC
[dev] Bug in Libcloud 0.20.0 GCE driver?
Hey Eric,
Are you aware of any bugs in the GCE driver in the latest Libcloud release
(0.20.0) - https://twitter.com/gadi_fe/status/686460086361952256
If so, and the fix is already in trunk, feel free to roll out 0.20.1
release with the bug fix. Keep in mind though, that you will probably need
to branch of v0.20.0 tag since a bunch of bigger changes have already
landed in trunk.
In any case, I can help with that if needed.
If there are no bugs and it's simply a lack of docs or similar, it would
also be great to address that :)
Thanks
Re: [dev] Bug in Libcloud 0.20.0 GCE driver?
Posted by Tomaz Muraus <to...@apache.org>.
I believe there were some some bug fixes in the Dimension Data drivers.
Anthony probably has a better idea which commits / PRs represent those
fixes.
On Thu, Jan 14, 2016 at 7:06 PM, Eric Johnson <er...@google.com.invalid>
wrote:
> From what I see, I'd say these are bugs fixes we should included in 0.20.1
>
> 4632db2d1ee5329ab2c7421c1c5c7ee1ef1e36b4
> 2e036dae4a41b59009eb2205df4d3f1ec03beee2
> f755d5597793bc8f3109f09e6b39c4ace2d740bf
> 3e85cf1591c9ccb07ec779ccb6a589e56534799d
>
> I'll stall here until later in the day to give others a chance to provide
> input on other cherry picks.
>
>
> On Thu, Jan 14, 2016 at 9:43 AM, Eric Johnson <er...@google.com> wrote:
>
> > Great. I'll skim over the log looking for bug fixes, but if anyone knows
> > of any, please flag them for me. :)
> >
> > On Thu, Jan 14, 2016 at 9:41 AM, Tomaz Muraus <to...@apache.org> wrote:
> >
> >> Yeah, you just need to make sure 0.20.1 branch is based off v0.20.0 tag.
> >>
> >> And to make it easier to differentiate between tags and branches, please
> >> drop the "v" prefix from the branch name (just 0.20.1 for branch name
> and
> >> v0.21.1 for tag).
> >>
> >> Then you cherry pick stuff (there might be conflicts so you might need
> to
> >> do some manual editing) and rest of the process is the same as usual.
> >>
> >> If there are any other bug fixes in trunk worth picking, we should also
> >> include those.
> >>
> >> On Thu, Jan 14, 2016 at 6:35 PM, Eric Johnson
> <erjohnso@google.com.invalid
> >> >
> >> wrote:
> >>
> >> > Ok, pretty sure this is the problem.
> >> >
> >> > Can you sanity check my process for creating the release artifacts?
> >> >
> >> > 1) make sure my local git repo trunk is up to date
> >> > 2) git checkout -b v0.20.1 4087f45504f7c3bbd0f389420e9ecbdedf93ded9 #
> >> hash
> >> > for 0.20.0
> >> > 3) git cherry-pick 88170f6582dd6cc3bce87c0ab3a89118551fe508
> >> #
> >> > hash for the GCE client id fix
> >> > 4) Test (see, I can learn from my mistakes!)
> >> > 5) bump __version__ in libcloud/__init__.py and commit # use this
> hash
> >> > below for -tentative tag
> >> > 6) git push origin v0.20.1
> >> > 7) git tag v0.20.1-tentative <hash_from_above>
> >> > 8) build and upload artifacts and start voting thread
> >> >
> >> > Sound ok?
> >> >
> >> >
> >> > On Wed, Jan 13, 2016 at 5:30 AM, Tomaz Muraus <to...@apache.org>
> wrote:
> >> >
> >> > > Yeah, need to spin up voting thread and follow the usual release
> >> process.
> >> > >
> >> > > Each release artifacts need to be trackable back to the git repo
> SHA,
> >> > etc.
> >> > >
> >> > > On Wed, Jan 13, 2016 at 2:06 PM, Eric Johnson
> >> > <erjohnso@google.com.invalid
> >> > > >
> >> > > wrote:
> >> > >
> >> > > > Hi Tomaz,
> >> > > >
> >> > > > I'll look more closely but I suspect the problem is that Google
> >> changed
> >> > > the
> >> > > > client id address format (and that is fixed in trunk).
> >> > > >
> >> > > > Do I need to spin up a vote thread to pull in that single fix?
> >> > > >
> >> > > > Thanks for flagging this!
> >> > > > Eric
> >> > > > On Jan 13, 2016 12:52 AM, "Tomaz Muraus" <to...@apache.org>
> wrote:
> >> > > >
> >> > > > > Hey Eric,
> >> > > > >
> >> > > > > Are you aware of any bugs in the GCE driver in the latest
> Libcloud
> >> > > > release
> >> > > > > (0.20.0) -
> https://twitter.com/gadi_fe/status/686460086361952256
> >> > > > >
> >> > > > > If so, and the fix is already in trunk, feel free to roll out
> >> 0.20.1
> >> > > > > release with the bug fix. Keep in mind though, that you will
> >> probably
> >> > > > need
> >> > > > > to branch of v0.20.0 tag since a bunch of bigger changes have
> >> already
> >> > > > > landed in trunk.
> >> > > > >
> >> > > > > In any case, I can help with that if needed.
> >> > > > >
> >> > > > > If there are no bugs and it's simply a lack of docs or similar,
> it
> >> > > would
> >> > > > > also be great to address that :)
> >> > > > >
> >> > > > > Thanks
> >> > > > >
> >> > > >
> >> > >
> >> >
> >>
> >
> >
>
Re: [dev] Bug in Libcloud 0.20.0 GCE driver?
Posted by Eric Johnson <er...@google.com.INVALID>.
>From what I see, I'd say these are bugs fixes we should included in 0.20.1
4632db2d1ee5329ab2c7421c1c5c7ee1ef1e36b4
2e036dae4a41b59009eb2205df4d3f1ec03beee2
f755d5597793bc8f3109f09e6b39c4ace2d740bf
3e85cf1591c9ccb07ec779ccb6a589e56534799d
I'll stall here until later in the day to give others a chance to provide
input on other cherry picks.
On Thu, Jan 14, 2016 at 9:43 AM, Eric Johnson <er...@google.com> wrote:
> Great. I'll skim over the log looking for bug fixes, but if anyone knows
> of any, please flag them for me. :)
>
> On Thu, Jan 14, 2016 at 9:41 AM, Tomaz Muraus <to...@apache.org> wrote:
>
>> Yeah, you just need to make sure 0.20.1 branch is based off v0.20.0 tag.
>>
>> And to make it easier to differentiate between tags and branches, please
>> drop the "v" prefix from the branch name (just 0.20.1 for branch name and
>> v0.21.1 for tag).
>>
>> Then you cherry pick stuff (there might be conflicts so you might need to
>> do some manual editing) and rest of the process is the same as usual.
>>
>> If there are any other bug fixes in trunk worth picking, we should also
>> include those.
>>
>> On Thu, Jan 14, 2016 at 6:35 PM, Eric Johnson <erjohnso@google.com.invalid
>> >
>> wrote:
>>
>> > Ok, pretty sure this is the problem.
>> >
>> > Can you sanity check my process for creating the release artifacts?
>> >
>> > 1) make sure my local git repo trunk is up to date
>> > 2) git checkout -b v0.20.1 4087f45504f7c3bbd0f389420e9ecbdedf93ded9 #
>> hash
>> > for 0.20.0
>> > 3) git cherry-pick 88170f6582dd6cc3bce87c0ab3a89118551fe508
>> #
>> > hash for the GCE client id fix
>> > 4) Test (see, I can learn from my mistakes!)
>> > 5) bump __version__ in libcloud/__init__.py and commit # use this hash
>> > below for -tentative tag
>> > 6) git push origin v0.20.1
>> > 7) git tag v0.20.1-tentative <hash_from_above>
>> > 8) build and upload artifacts and start voting thread
>> >
>> > Sound ok?
>> >
>> >
>> > On Wed, Jan 13, 2016 at 5:30 AM, Tomaz Muraus <to...@apache.org> wrote:
>> >
>> > > Yeah, need to spin up voting thread and follow the usual release
>> process.
>> > >
>> > > Each release artifacts need to be trackable back to the git repo SHA,
>> > etc.
>> > >
>> > > On Wed, Jan 13, 2016 at 2:06 PM, Eric Johnson
>> > <erjohnso@google.com.invalid
>> > > >
>> > > wrote:
>> > >
>> > > > Hi Tomaz,
>> > > >
>> > > > I'll look more closely but I suspect the problem is that Google
>> changed
>> > > the
>> > > > client id address format (and that is fixed in trunk).
>> > > >
>> > > > Do I need to spin up a vote thread to pull in that single fix?
>> > > >
>> > > > Thanks for flagging this!
>> > > > Eric
>> > > > On Jan 13, 2016 12:52 AM, "Tomaz Muraus" <to...@apache.org> wrote:
>> > > >
>> > > > > Hey Eric,
>> > > > >
>> > > > > Are you aware of any bugs in the GCE driver in the latest Libcloud
>> > > > release
>> > > > > (0.20.0) - https://twitter.com/gadi_fe/status/686460086361952256
>> > > > >
>> > > > > If so, and the fix is already in trunk, feel free to roll out
>> 0.20.1
>> > > > > release with the bug fix. Keep in mind though, that you will
>> probably
>> > > > need
>> > > > > to branch of v0.20.0 tag since a bunch of bigger changes have
>> already
>> > > > > landed in trunk.
>> > > > >
>> > > > > In any case, I can help with that if needed.
>> > > > >
>> > > > > If there are no bugs and it's simply a lack of docs or similar, it
>> > > would
>> > > > > also be great to address that :)
>> > > > >
>> > > > > Thanks
>> > > > >
>> > > >
>> > >
>> >
>>
>
>
Re: [dev] Bug in Libcloud 0.20.0 GCE driver?
Posted by Eric Johnson <er...@google.com.INVALID>.
Great. I'll skim over the log looking for bug fixes, but if anyone knows
of any, please flag them for me. :)
On Thu, Jan 14, 2016 at 9:41 AM, Tomaz Muraus <to...@apache.org> wrote:
> Yeah, you just need to make sure 0.20.1 branch is based off v0.20.0 tag.
>
> And to make it easier to differentiate between tags and branches, please
> drop the "v" prefix from the branch name (just 0.20.1 for branch name and
> v0.21.1 for tag).
>
> Then you cherry pick stuff (there might be conflicts so you might need to
> do some manual editing) and rest of the process is the same as usual.
>
> If there are any other bug fixes in trunk worth picking, we should also
> include those.
>
> On Thu, Jan 14, 2016 at 6:35 PM, Eric Johnson <erjohnso@google.com.invalid
> >
> wrote:
>
> > Ok, pretty sure this is the problem.
> >
> > Can you sanity check my process for creating the release artifacts?
> >
> > 1) make sure my local git repo trunk is up to date
> > 2) git checkout -b v0.20.1 4087f45504f7c3bbd0f389420e9ecbdedf93ded9 #
> hash
> > for 0.20.0
> > 3) git cherry-pick 88170f6582dd6cc3bce87c0ab3a89118551fe508
> #
> > hash for the GCE client id fix
> > 4) Test (see, I can learn from my mistakes!)
> > 5) bump __version__ in libcloud/__init__.py and commit # use this hash
> > below for -tentative tag
> > 6) git push origin v0.20.1
> > 7) git tag v0.20.1-tentative <hash_from_above>
> > 8) build and upload artifacts and start voting thread
> >
> > Sound ok?
> >
> >
> > On Wed, Jan 13, 2016 at 5:30 AM, Tomaz Muraus <to...@apache.org> wrote:
> >
> > > Yeah, need to spin up voting thread and follow the usual release
> process.
> > >
> > > Each release artifacts need to be trackable back to the git repo SHA,
> > etc.
> > >
> > > On Wed, Jan 13, 2016 at 2:06 PM, Eric Johnson
> > <erjohnso@google.com.invalid
> > > >
> > > wrote:
> > >
> > > > Hi Tomaz,
> > > >
> > > > I'll look more closely but I suspect the problem is that Google
> changed
> > > the
> > > > client id address format (and that is fixed in trunk).
> > > >
> > > > Do I need to spin up a vote thread to pull in that single fix?
> > > >
> > > > Thanks for flagging this!
> > > > Eric
> > > > On Jan 13, 2016 12:52 AM, "Tomaz Muraus" <to...@apache.org> wrote:
> > > >
> > > > > Hey Eric,
> > > > >
> > > > > Are you aware of any bugs in the GCE driver in the latest Libcloud
> > > > release
> > > > > (0.20.0) - https://twitter.com/gadi_fe/status/686460086361952256
> > > > >
> > > > > If so, and the fix is already in trunk, feel free to roll out
> 0.20.1
> > > > > release with the bug fix. Keep in mind though, that you will
> probably
> > > > need
> > > > > to branch of v0.20.0 tag since a bunch of bigger changes have
> already
> > > > > landed in trunk.
> > > > >
> > > > > In any case, I can help with that if needed.
> > > > >
> > > > > If there are no bugs and it's simply a lack of docs or similar, it
> > > would
> > > > > also be great to address that :)
> > > > >
> > > > > Thanks
> > > > >
> > > >
> > >
> >
>
Re: [dev] Bug in Libcloud 0.20.0 GCE driver?
Posted by Tomaz Muraus <to...@apache.org>.
Yeah, you just need to make sure 0.20.1 branch is based off v0.20.0 tag.
And to make it easier to differentiate between tags and branches, please
drop the "v" prefix from the branch name (just 0.20.1 for branch name and
v0.21.1 for tag).
Then you cherry pick stuff (there might be conflicts so you might need to
do some manual editing) and rest of the process is the same as usual.
If there are any other bug fixes in trunk worth picking, we should also
include those.
On Thu, Jan 14, 2016 at 6:35 PM, Eric Johnson <er...@google.com.invalid>
wrote:
> Ok, pretty sure this is the problem.
>
> Can you sanity check my process for creating the release artifacts?
>
> 1) make sure my local git repo trunk is up to date
> 2) git checkout -b v0.20.1 4087f45504f7c3bbd0f389420e9ecbdedf93ded9 # hash
> for 0.20.0
> 3) git cherry-pick 88170f6582dd6cc3bce87c0ab3a89118551fe508 #
> hash for the GCE client id fix
> 4) Test (see, I can learn from my mistakes!)
> 5) bump __version__ in libcloud/__init__.py and commit # use this hash
> below for -tentative tag
> 6) git push origin v0.20.1
> 7) git tag v0.20.1-tentative <hash_from_above>
> 8) build and upload artifacts and start voting thread
>
> Sound ok?
>
>
> On Wed, Jan 13, 2016 at 5:30 AM, Tomaz Muraus <to...@apache.org> wrote:
>
> > Yeah, need to spin up voting thread and follow the usual release process.
> >
> > Each release artifacts need to be trackable back to the git repo SHA,
> etc.
> >
> > On Wed, Jan 13, 2016 at 2:06 PM, Eric Johnson
> <erjohnso@google.com.invalid
> > >
> > wrote:
> >
> > > Hi Tomaz,
> > >
> > > I'll look more closely but I suspect the problem is that Google changed
> > the
> > > client id address format (and that is fixed in trunk).
> > >
> > > Do I need to spin up a vote thread to pull in that single fix?
> > >
> > > Thanks for flagging this!
> > > Eric
> > > On Jan 13, 2016 12:52 AM, "Tomaz Muraus" <to...@apache.org> wrote:
> > >
> > > > Hey Eric,
> > > >
> > > > Are you aware of any bugs in the GCE driver in the latest Libcloud
> > > release
> > > > (0.20.0) - https://twitter.com/gadi_fe/status/686460086361952256
> > > >
> > > > If so, and the fix is already in trunk, feel free to roll out 0.20.1
> > > > release with the bug fix. Keep in mind though, that you will probably
> > > need
> > > > to branch of v0.20.0 tag since a bunch of bigger changes have already
> > > > landed in trunk.
> > > >
> > > > In any case, I can help with that if needed.
> > > >
> > > > If there are no bugs and it's simply a lack of docs or similar, it
> > would
> > > > also be great to address that :)
> > > >
> > > > Thanks
> > > >
> > >
> >
>
Re: [dev] Bug in Libcloud 0.20.0 GCE driver?
Posted by Eric Johnson <er...@google.com.INVALID>.
Ok, pretty sure this is the problem.
Can you sanity check my process for creating the release artifacts?
1) make sure my local git repo trunk is up to date
2) git checkout -b v0.20.1 4087f45504f7c3bbd0f389420e9ecbdedf93ded9 # hash
for 0.20.0
3) git cherry-pick 88170f6582dd6cc3bce87c0ab3a89118551fe508 #
hash for the GCE client id fix
4) Test (see, I can learn from my mistakes!)
5) bump __version__ in libcloud/__init__.py and commit # use this hash
below for -tentative tag
6) git push origin v0.20.1
7) git tag v0.20.1-tentative <hash_from_above>
8) build and upload artifacts and start voting thread
Sound ok?
On Wed, Jan 13, 2016 at 5:30 AM, Tomaz Muraus <to...@apache.org> wrote:
> Yeah, need to spin up voting thread and follow the usual release process.
>
> Each release artifacts need to be trackable back to the git repo SHA, etc.
>
> On Wed, Jan 13, 2016 at 2:06 PM, Eric Johnson <erjohnso@google.com.invalid
> >
> wrote:
>
> > Hi Tomaz,
> >
> > I'll look more closely but I suspect the problem is that Google changed
> the
> > client id address format (and that is fixed in trunk).
> >
> > Do I need to spin up a vote thread to pull in that single fix?
> >
> > Thanks for flagging this!
> > Eric
> > On Jan 13, 2016 12:52 AM, "Tomaz Muraus" <to...@apache.org> wrote:
> >
> > > Hey Eric,
> > >
> > > Are you aware of any bugs in the GCE driver in the latest Libcloud
> > release
> > > (0.20.0) - https://twitter.com/gadi_fe/status/686460086361952256
> > >
> > > If so, and the fix is already in trunk, feel free to roll out 0.20.1
> > > release with the bug fix. Keep in mind though, that you will probably
> > need
> > > to branch of v0.20.0 tag since a bunch of bigger changes have already
> > > landed in trunk.
> > >
> > > In any case, I can help with that if needed.
> > >
> > > If there are no bugs and it's simply a lack of docs or similar, it
> would
> > > also be great to address that :)
> > >
> > > Thanks
> > >
> >
>
Re: [dev] Bug in Libcloud 0.20.0 GCE driver?
Posted by Tomaz Muraus <to...@apache.org>.
Yeah, need to spin up voting thread and follow the usual release process.
Each release artifacts need to be trackable back to the git repo SHA, etc.
On Wed, Jan 13, 2016 at 2:06 PM, Eric Johnson <er...@google.com.invalid>
wrote:
> Hi Tomaz,
>
> I'll look more closely but I suspect the problem is that Google changed the
> client id address format (and that is fixed in trunk).
>
> Do I need to spin up a vote thread to pull in that single fix?
>
> Thanks for flagging this!
> Eric
> On Jan 13, 2016 12:52 AM, "Tomaz Muraus" <to...@apache.org> wrote:
>
> > Hey Eric,
> >
> > Are you aware of any bugs in the GCE driver in the latest Libcloud
> release
> > (0.20.0) - https://twitter.com/gadi_fe/status/686460086361952256
> >
> > If so, and the fix is already in trunk, feel free to roll out 0.20.1
> > release with the bug fix. Keep in mind though, that you will probably
> need
> > to branch of v0.20.0 tag since a bunch of bigger changes have already
> > landed in trunk.
> >
> > In any case, I can help with that if needed.
> >
> > If there are no bugs and it's simply a lack of docs or similar, it would
> > also be great to address that :)
> >
> > Thanks
> >
>
Re: [dev] Bug in Libcloud 0.20.0 GCE driver?
Posted by Eric Johnson <er...@google.com.INVALID>.
Hi Tomaz,
I'll look more closely but I suspect the problem is that Google changed the
client id address format (and that is fixed in trunk).
Do I need to spin up a vote thread to pull in that single fix?
Thanks for flagging this!
Eric
On Jan 13, 2016 12:52 AM, "Tomaz Muraus" <to...@apache.org> wrote:
> Hey Eric,
>
> Are you aware of any bugs in the GCE driver in the latest Libcloud release
> (0.20.0) - https://twitter.com/gadi_fe/status/686460086361952256
>
> If so, and the fix is already in trunk, feel free to roll out 0.20.1
> release with the bug fix. Keep in mind though, that you will probably need
> to branch of v0.20.0 tag since a bunch of bigger changes have already
> landed in trunk.
>
> In any case, I can help with that if needed.
>
> If there are no bugs and it's simply a lack of docs or similar, it would
> also be great to address that :)
>
> Thanks
>