[infra][release] delete old EOL'd stable branches
Hi Infra Team! In October there was a discussion at Release Team meeting [1] about what can we do with the old, already EOL'd but not yet deleted branches (this is possible since with the Extended Maintenance process the general/"mass" EOL'ing was stopped and tagging a project branch EOL does not delete the branch anymore). Related to this, I would like to ask two things: 1. I've used the list_eol_stale_branches.sh [2] script to get the list of such not-yet-deleted branches for Ocata [3]. They are all tagged with 'ocata-eol', but stable/ocata branch still exists for them. Could you please delete these? [3] 2. On the Release Team meeting [1] we were hinted that with the newer version of gerrit (that was installed at the end of November) some automation is possible through gerrit API in the future. Can I get some help about where should I start with the automation? Which repository should I look, where can the deletion being triggered ("similarly like branch creation")? Thanks in advance, Előd [1] http://eavesdrop.openstack.org/meetings/releaseteam/2020/releaseteam.2020-10... [2] https://opendev.org/openstack/releases/src/commit/eb381492da3f7c826c35b9f147... [3] http://paste.openstack.org/show/801992/
On 2021-01-26 17:52:07 +0100 (+0100), Előd Illés wrote: [...]
1. I've used the list_eol_stale_branches.sh [2] script to get the list of such not-yet-deleted branches for Ocata [3]. They are all tagged with 'ocata-eol', but stable/ocata branch still exists for them. Could you please delete these? [3]
I'm happy to, have you made sure any open reviews for those branches are abandoned first? Gerrit won't allow deletion of a branch with open reviews.
2. On the Release Team meeting [1] we were hinted that with the newer version of gerrit (that was installed at the end of November) some automation is possible through gerrit API in the future. Can I get some help about where should I start with the automation? Which repository should I look, where can the deletion being triggered ("similarly like branch creation")? [...]
The Gerrit REST API method for deleting branches is documented here: https://review.opendev.org/Documentation/rest-api-projects.html#delete-branc... I'm not immediately sure where branch creation happens in the forest of our release automation, but I would expect deletion could be implemented similarly. Hopefully someone more intimately familiar with those jobs can chime in. The access control we'll need to grant to automation so that it can call that is documented here: https://review.opendev.org/Documentation/access-control.html#category_delete It'll need to be added manually as a permission for the Release Managers group in our All-Projects global ACL which individual projects inherit, and this documentation updated accordingly: https://opendev.org/opendev/system-config/src/branch/master/doc/source/gerri... Happy to answer other questions as they arise. -- Jeremy Stanley
Le mar. 26 janv. 2021 à 18:35, Jeremy Stanley <fungi@yuggoth.org> a écrit :
1. I've used the list_eol_stale_branches.sh [2] script to get the list of such not-yet-deleted branches for Ocata [3]. They are all tagged with 'ocata-eol', but stable/ocata branch still exists for them. Could you
On 2021-01-26 17:52:07 +0100 (+0100), Előd Illés wrote: [...] please
delete these? [3]
I'm happy to, have you made sure any open reviews for those branches are abandoned first? Gerrit won't allow deletion of a branch with open reviews.
I think we need a first round of inspection on these stale branches to see if opened patches exist and then if needed start discussion with teams to ask them to drop the patches those who have been found. I'll try to add this feature to check for opened patches within `list_eol_stale_branches.sh`.
2. On the Release Team meeting [1] we were hinted that with the newer version of gerrit (that was installed at the end of November) some automation is possible through gerrit API in the future. Can I get some help about where should I start with the automation? Which repository should I look, where can the deletion being triggered ("similarly like branch creation")? [...]
The Gerrit REST API method for deleting branches is documented here:
https://review.opendev.org/Documentation/rest-api-projects.html#delete-branc...
I'm not immediately sure where branch creation happens in the forest of our release automation, but I would expect deletion could be implemented similarly. Hopefully someone more intimately familiar with those jobs can chime in.
The access control we'll need to grant to automation so that it can call that is documented here:
https://review.opendev.org/Documentation/access-control.html#category_delete
It'll need to be added manually as a permission for the Release Managers group in our All-Projects global ACL which individual projects inherit, and this documentation updated accordingly:
https://opendev.org/opendev/system-config/src/branch/master/doc/source/gerri...
Happy to answer other questions as they arise. -- Jeremy Stanley
-- Hervé Beraud Senior Software Engineer at Red Hat irc: hberaud https://github.com/4383/ https://twitter.com/4383hberaud -----BEGIN PGP SIGNATURE----- wsFcBAABCAAQBQJb4AwCCRAHwXRBNkGNegAALSkQAHrotwCiL3VMwDR0vcja10Q+ Kf31yCutl5bAlS7tOKpPQ9XN4oC0ZSThyNNFVrg8ail0SczHXsC4rOrsPblgGRN+ RQLoCm2eO1AkB0ubCYLaq0XqSaO+Uk81QxAPkyPCEGT6SRxXr2lhADK0T86kBnMP F8RvGolu3EFjlqCVgeOZaR51PqwUlEhZXZuuNKrWZXg/oRiY4811GmnvzmUhgK5G 5+f8mUg74hfjDbR2VhjTeaLKp0PhskjOIKY3vqHXofLuaqFDD+WrAy/NgDGvN22g glGfj472T3xyHnUzM8ILgAGSghfzZF5Skj2qEeci9cB6K3Hm3osj+PbvfsXE/7Kw m/xtm+FjnaywZEv54uCmVIzQsRIm1qJscu20Qw6Q0UiPpDFqD7O6tWSRKdX11UTZ hwVQTMh9AKQDBEh2W9nnFi9kzSSNu4OQ1dRMcYHWfd9BEkccezxHwUM4Xyov5Fe0 qnbfzTB1tYkjU78loMWFaLa00ftSxP/DtQ//iYVyfVNfcCwfDszXLOqlkvGmY1/Y F1ON0ONekDZkGJsDoS6QdiUSn8RZ2mHArGEWMV00EV5DCIbCXRvywXV43ckx8Z+3 B8qUJhBqJ8RS2F+vTs3DTaXqcktgJ4UkhYC2c1gImcPRyGrK9VY0sCT+1iA+wp/O v6rDpkeNksZ9fFSyoY2o =ECSj -----END PGP SIGNATURE-----
Thanks Jeremy and Hervé! @Jeremy, @Hervé: I've checked and found two open patches, out of which I was able to abandon one, and asked the owner on the other patch to abandon it. So the listed branches can be deleted *except* one: openstack/os-collect-config So in my opinion there's no need for further coordination with the teams, as these branches are tagged ocata-eol already. And thanks Hervé for the script, it helped a lot so far, already :) + thanks Jeremy for the pointers! Előd On 2021. 01. 27. 11:23, Herve Beraud wrote:
Le mar. 26 janv. 2021 à 18:35, Jeremy Stanley <fungi@yuggoth.org <mailto:fungi@yuggoth.org>> a écrit :
On 2021-01-26 17:52:07 +0100 (+0100), Előd Illés wrote: [...] > 1. I've used the list_eol_stale_branches.sh [2] script to get the list of > such not-yet-deleted branches for Ocata [3]. They are all tagged with > 'ocata-eol', but stable/ocata branch still exists for them. Could you please > delete these? [3]
I'm happy to, have you made sure any open reviews for those branches are abandoned first? Gerrit won't allow deletion of a branch with open reviews.
I think we need a first round of inspection on these stale branches to see if opened patches exist and then if needed start discussion with teams to ask them to drop the patches those who have been found.
I'll try to add this feature to check for opened patches within `list_eol_stale_branches.sh`.
> 2. On the Release Team meeting [1] we were hinted that with the newer > version of gerrit (that was installed at the end of November) some > automation is possible through gerrit API in the future. Can I get some help > about where should I start with the automation? Which repository should I > look, where can the deletion being triggered ("similarly like branch > creation")? [...]
The Gerrit REST API method for deleting branches is documented here:
https://review.opendev.org/Documentation/rest-api-projects.html#delete-branc...
I'm not immediately sure where branch creation happens in the forest of our release automation, but I would expect deletion could be implemented similarly. Hopefully someone more intimately familiar with those jobs can chime in.
The access control we'll need to grant to automation so that it can call that is documented here:
https://review.opendev.org/Documentation/access-control.html#category_delete
It'll need to be added manually as a permission for the Release Managers group in our All-Projects global ACL which individual projects inherit, and this documentation updated accordingly:
https://opendev.org/opendev/system-config/src/branch/master/doc/source/gerri...
Happy to answer other questions as they arise. -- Jeremy Stanley
-- Hervé Beraud Senior Software Engineer at Red Hat irc: hberaud https://github.com/4383/ https://twitter.com/4383hberaud -----BEGIN PGP SIGNATURE-----
wsFcBAABCAAQBQJb4AwCCRAHwXRBNkGNegAALSkQAHrotwCiL3VMwDR0vcja10Q+ Kf31yCutl5bAlS7tOKpPQ9XN4oC0ZSThyNNFVrg8ail0SczHXsC4rOrsPblgGRN+ RQLoCm2eO1AkB0ubCYLaq0XqSaO+Uk81QxAPkyPCEGT6SRxXr2lhADK0T86kBnMP F8RvGolu3EFjlqCVgeOZaR51PqwUlEhZXZuuNKrWZXg/oRiY4811GmnvzmUhgK5G 5+f8mUg74hfjDbR2VhjTeaLKp0PhskjOIKY3vqHXofLuaqFDD+WrAy/NgDGvN22g glGfj472T3xyHnUzM8ILgAGSghfzZF5Skj2qEeci9cB6K3Hm3osj+PbvfsXE/7Kw m/xtm+FjnaywZEv54uCmVIzQsRIm1qJscu20Qw6Q0UiPpDFqD7O6tWSRKdX11UTZ hwVQTMh9AKQDBEh2W9nnFi9kzSSNu4OQ1dRMcYHWfd9BEkccezxHwUM4Xyov5Fe0 qnbfzTB1tYkjU78loMWFaLa00ftSxP/DtQ//iYVyfVNfcCwfDszXLOqlkvGmY1/Y F1ON0ONekDZkGJsDoS6QdiUSn8RZ2mHArGEWMV00EV5DCIbCXRvywXV43ckx8Z+3 B8qUJhBqJ8RS2F+vTs3DTaXqcktgJ4UkhYC2c1gImcPRyGrK9VY0sCT+1iA+wp/O v6rDpkeNksZ9fFSyoY2o =ECSj -----END PGP SIGNATURE-----
Le mer. 27 janv. 2021 à 11:51, Előd Illés <elod.illes@est.tech> a écrit :
Thanks Jeremy and Hervé!
@Jeremy, @Hervé: I've checked and found two open patches, out of which I was able to abandon one, and asked the owner on the other patch to abandon it. So the listed branches can be deleted *except* one: openstack/os-collect-config
So in my opinion there's no need for further coordination with the teams, as these branches are tagged ocata-eol already.
Yes indeed, I agree.
And thanks Hervé for the script, it helped a lot so far, already :) + thanks Jeremy for the pointers!
Előd
On 2021. 01. 27. 11:23, Herve Beraud wrote:
Le mar. 26 janv. 2021 à 18:35, Jeremy Stanley <fungi@yuggoth.org> a écrit :
1. I've used the list_eol_stale_branches.sh [2] script to get the list of such not-yet-deleted branches for Ocata [3]. They are all tagged with 'ocata-eol', but stable/ocata branch still exists for them. Could you
On 2021-01-26 17:52:07 +0100 (+0100), Előd Illés wrote: [...] please
delete these? [3]
I'm happy to, have you made sure any open reviews for those branches are abandoned first? Gerrit won't allow deletion of a branch with open reviews.
I think we need a first round of inspection on these stale branches to see if opened patches exist and then if needed start discussion with teams to ask them to drop the patches those who have been found.
I'll try to add this feature to check for opened patches within `list_eol_stale_branches.sh`.
2. On the Release Team meeting [1] we were hinted that with the newer version of gerrit (that was installed at the end of November) some automation is possible through gerrit API in the future. Can I get some help about where should I start with the automation? Which repository should I look, where can the deletion being triggered ("similarly like branch creation")? [...]
The Gerrit REST API method for deleting branches is documented here:
https://review.opendev.org/Documentation/rest-api-projects.html#delete-branc...
I'm not immediately sure where branch creation happens in the forest of our release automation, but I would expect deletion could be implemented similarly. Hopefully someone more intimately familiar with those jobs can chime in.
The access control we'll need to grant to automation so that it can call that is documented here:
https://review.opendev.org/Documentation/access-control.html#category_delete
It'll need to be added manually as a permission for the Release Managers group in our All-Projects global ACL which individual projects inherit, and this documentation updated accordingly:
https://opendev.org/opendev/system-config/src/branch/master/doc/source/gerri...
Happy to answer other questions as they arise. -- Jeremy Stanley
-- Hervé Beraud Senior Software Engineer at Red Hat irc: hberaud https://github.com/4383/ https://twitter.com/4383hberaud -----BEGIN PGP SIGNATURE-----
wsFcBAABCAAQBQJb4AwCCRAHwXRBNkGNegAALSkQAHrotwCiL3VMwDR0vcja10Q+ Kf31yCutl5bAlS7tOKpPQ9XN4oC0ZSThyNNFVrg8ail0SczHXsC4rOrsPblgGRN+ RQLoCm2eO1AkB0ubCYLaq0XqSaO+Uk81QxAPkyPCEGT6SRxXr2lhADK0T86kBnMP F8RvGolu3EFjlqCVgeOZaR51PqwUlEhZXZuuNKrWZXg/oRiY4811GmnvzmUhgK5G 5+f8mUg74hfjDbR2VhjTeaLKp0PhskjOIKY3vqHXofLuaqFDD+WrAy/NgDGvN22g glGfj472T3xyHnUzM8ILgAGSghfzZF5Skj2qEeci9cB6K3Hm3osj+PbvfsXE/7Kw m/xtm+FjnaywZEv54uCmVIzQsRIm1qJscu20Qw6Q0UiPpDFqD7O6tWSRKdX11UTZ hwVQTMh9AKQDBEh2W9nnFi9kzSSNu4OQ1dRMcYHWfd9BEkccezxHwUM4Xyov5Fe0 qnbfzTB1tYkjU78loMWFaLa00ftSxP/DtQ//iYVyfVNfcCwfDszXLOqlkvGmY1/Y F1ON0ONekDZkGJsDoS6QdiUSn8RZ2mHArGEWMV00EV5DCIbCXRvywXV43ckx8Z+3 B8qUJhBqJ8RS2F+vTs3DTaXqcktgJ4UkhYC2c1gImcPRyGrK9VY0sCT+1iA+wp/O v6rDpkeNksZ9fFSyoY2o =ECSj -----END PGP SIGNATURE-----
-- Hervé Beraud Senior Software Engineer at Red Hat irc: hberaud https://github.com/4383/ https://twitter.com/4383hberaud -----BEGIN PGP SIGNATURE----- wsFcBAABCAAQBQJb4AwCCRAHwXRBNkGNegAALSkQAHrotwCiL3VMwDR0vcja10Q+ Kf31yCutl5bAlS7tOKpPQ9XN4oC0ZSThyNNFVrg8ail0SczHXsC4rOrsPblgGRN+ RQLoCm2eO1AkB0ubCYLaq0XqSaO+Uk81QxAPkyPCEGT6SRxXr2lhADK0T86kBnMP F8RvGolu3EFjlqCVgeOZaR51PqwUlEhZXZuuNKrWZXg/oRiY4811GmnvzmUhgK5G 5+f8mUg74hfjDbR2VhjTeaLKp0PhskjOIKY3vqHXofLuaqFDD+WrAy/NgDGvN22g glGfj472T3xyHnUzM8ILgAGSghfzZF5Skj2qEeci9cB6K3Hm3osj+PbvfsXE/7Kw m/xtm+FjnaywZEv54uCmVIzQsRIm1qJscu20Qw6Q0UiPpDFqD7O6tWSRKdX11UTZ hwVQTMh9AKQDBEh2W9nnFi9kzSSNu4OQ1dRMcYHWfd9BEkccezxHwUM4Xyov5Fe0 qnbfzTB1tYkjU78loMWFaLa00ftSxP/DtQ//iYVyfVNfcCwfDszXLOqlkvGmY1/Y F1ON0ONekDZkGJsDoS6QdiUSn8RZ2mHArGEWMV00EV5DCIbCXRvywXV43ckx8Z+3 B8qUJhBqJ8RS2F+vTs3DTaXqcktgJ4UkhYC2c1gImcPRyGrK9VY0sCT+1iA+wp/O v6rDpkeNksZ9fFSyoY2o =ECSj -----END PGP SIGNATURE-----
Le mer. 27 janv. 2021 à 11:23, Herve Beraud <hberaud@redhat.com> a écrit :
Le mar. 26 janv. 2021 à 18:35, Jeremy Stanley <fungi@yuggoth.org> a écrit :
1. I've used the list_eol_stale_branches.sh [2] script to get the list of such not-yet-deleted branches for Ocata [3]. They are all tagged with 'ocata-eol', but stable/ocata branch still exists for them. Could you
On 2021-01-26 17:52:07 +0100 (+0100), Előd Illés wrote: [...] please
delete these? [3]
I'm happy to, have you made sure any open reviews for those branches are abandoned first? Gerrit won't allow deletion of a branch with open reviews.
I think we need a first round of inspection on these stale branches to see if opened patches exist and then if needed start discussion with teams to ask them to drop the patches those who have been found.
I'll try to add this feature to check for opened patches within `list_eol_stale_branches.sh`.
I created a quick and dirty script [1] to inspect all these repos branches (based on the list given previously) and only os-collect-config still contains unmerged patches [2]. [1] http://paste.openstack.org/show/802035/ [2] https://review.opendev.org/q/project:openstack/os-collect-config+branch:stab...
2. On the Release Team meeting [1] we were hinted that with the newer version of gerrit (that was installed at the end of November) some automation is possible through gerrit API in the future. Can I get some help about where should I start with the automation? Which repository should I look, where can the deletion being triggered ("similarly like branch creation")? [...]
The Gerrit REST API method for deleting branches is documented here:
https://review.opendev.org/Documentation/rest-api-projects.html#delete-branc...
I'm not immediately sure where branch creation happens in the forest of our release automation, but I would expect deletion could be implemented similarly. Hopefully someone more intimately familiar with those jobs can chime in.
The access control we'll need to grant to automation so that it can call that is documented here:
https://review.opendev.org/Documentation/access-control.html#category_delete
It'll need to be added manually as a permission for the Release Managers group in our All-Projects global ACL which individual projects inherit, and this documentation updated accordingly:
https://opendev.org/opendev/system-config/src/branch/master/doc/source/gerri...
Happy to answer other questions as they arise. -- Jeremy Stanley
-- Hervé Beraud Senior Software Engineer at Red Hat irc: hberaud https://github.com/4383/ https://twitter.com/4383hberaud -----BEGIN PGP SIGNATURE-----
wsFcBAABCAAQBQJb4AwCCRAHwXRBNkGNegAALSkQAHrotwCiL3VMwDR0vcja10Q+ Kf31yCutl5bAlS7tOKpPQ9XN4oC0ZSThyNNFVrg8ail0SczHXsC4rOrsPblgGRN+ RQLoCm2eO1AkB0ubCYLaq0XqSaO+Uk81QxAPkyPCEGT6SRxXr2lhADK0T86kBnMP F8RvGolu3EFjlqCVgeOZaR51PqwUlEhZXZuuNKrWZXg/oRiY4811GmnvzmUhgK5G 5+f8mUg74hfjDbR2VhjTeaLKp0PhskjOIKY3vqHXofLuaqFDD+WrAy/NgDGvN22g glGfj472T3xyHnUzM8ILgAGSghfzZF5Skj2qEeci9cB6K3Hm3osj+PbvfsXE/7Kw m/xtm+FjnaywZEv54uCmVIzQsRIm1qJscu20Qw6Q0UiPpDFqD7O6tWSRKdX11UTZ hwVQTMh9AKQDBEh2W9nnFi9kzSSNu4OQ1dRMcYHWfd9BEkccezxHwUM4Xyov5Fe0 qnbfzTB1tYkjU78loMWFaLa00ftSxP/DtQ//iYVyfVNfcCwfDszXLOqlkvGmY1/Y F1ON0ONekDZkGJsDoS6QdiUSn8RZ2mHArGEWMV00EV5DCIbCXRvywXV43ckx8Z+3 B8qUJhBqJ8RS2F+vTs3DTaXqcktgJ4UkhYC2c1gImcPRyGrK9VY0sCT+1iA+wp/O v6rDpkeNksZ9fFSyoY2o =ECSj -----END PGP SIGNATURE-----
-- Hervé Beraud Senior Software Engineer at Red Hat irc: hberaud https://github.com/4383/ https://twitter.com/4383hberaud -----BEGIN PGP SIGNATURE----- wsFcBAABCAAQBQJb4AwCCRAHwXRBNkGNegAALSkQAHrotwCiL3VMwDR0vcja10Q+ Kf31yCutl5bAlS7tOKpPQ9XN4oC0ZSThyNNFVrg8ail0SczHXsC4rOrsPblgGRN+ RQLoCm2eO1AkB0ubCYLaq0XqSaO+Uk81QxAPkyPCEGT6SRxXr2lhADK0T86kBnMP F8RvGolu3EFjlqCVgeOZaR51PqwUlEhZXZuuNKrWZXg/oRiY4811GmnvzmUhgK5G 5+f8mUg74hfjDbR2VhjTeaLKp0PhskjOIKY3vqHXofLuaqFDD+WrAy/NgDGvN22g glGfj472T3xyHnUzM8ILgAGSghfzZF5Skj2qEeci9cB6K3Hm3osj+PbvfsXE/7Kw m/xtm+FjnaywZEv54uCmVIzQsRIm1qJscu20Qw6Q0UiPpDFqD7O6tWSRKdX11UTZ hwVQTMh9AKQDBEh2W9nnFi9kzSSNu4OQ1dRMcYHWfd9BEkccezxHwUM4Xyov5Fe0 qnbfzTB1tYkjU78loMWFaLa00ftSxP/DtQ//iYVyfVNfcCwfDszXLOqlkvGmY1/Y F1ON0ONekDZkGJsDoS6QdiUSn8RZ2mHArGEWMV00EV5DCIbCXRvywXV43ckx8Z+3 B8qUJhBqJ8RS2F+vTs3DTaXqcktgJ4UkhYC2c1gImcPRyGrK9VY0sCT+1iA+wp/O v6rDpkeNksZ9fFSyoY2o =ECSj -----END PGP SIGNATURE-----
On Tue, 2021-01-26 at 17:52 +0100, Előd Illés wrote: Hi Infra Team! In October there was a discussion at Release Team meeting [1] about what can we do with the old, already EOL'd but not yet deleted branches (this is possible since with the Extended Maintenance process the general/"mass" EOL'ing was stopped and tagging a project branch EOL does not delete the branch anymore). Not an answer but rather a question for my own understanding: what is the advantage of deleting branches? I understand that these things would no longer maintained and the gates will slowly break, but they're still relatively useful as a reference to explore project history and it's not like branches are expensive in git. Stephen Related to this, I would like to ask two things: 1. I've used the list_eol_stale_branches.sh [2] script to get the list of such not-yet-deleted branches for Ocata [3]. They are all tagged with 'ocata-eol', but stable/ocata branch still exists for them. Could you please delete these? [3] 2. On the Release Team meeting [1] we were hinted that with the newer version of gerrit (that was installed at the end of November) some automation is possible through gerrit API in the future. Can I get some help about where should I start with the automation? Which repository should I look, where can the deletion being triggered ("similarly like branch creation")? Thanks in advance, Előd [1] http://eavesdrop.openstack.org/meetings/releaseteam/2020/releaseteam.2020-10... [2] https://opendev.org/openstack/releases/src/commit/eb381492da3f7c826c35b9f147... [3] http://paste.openstack.org/show/801992/
Le mer. 27 janv. 2021 à 10:55, Stephen Finucane <stephenfin@redhat.com> a écrit :
On Tue, 2021-01-26 at 17:52 +0100, Előd Illés wrote: Hi Infra Team!
In October there was a discussion at Release Team meeting [1] about what can we do with the old, already EOL'd but not yet deleted branches (this is possible since with the Extended Maintenance process the general/"mass" EOL'ing was stopped and tagging a project branch EOL does not delete the branch anymore).
Not an answer but rather a question for my own understanding: what is the advantage of deleting branches? I understand that these things would no longer maintained and the gates will slowly break,
AFAIK this is mostly to avoid issues with gates/zuul.
but they're still relatively useful as a reference to explore project history and it's not like branches are expensive in git.
Tags can be used to dig in the related history.
Stephen
Related to this, I would like to ask two things:
1. I've used the list_eol_stale_branches.sh [2] script to get the list of such not-yet-deleted branches for Ocata [3]. They are all tagged with 'ocata-eol', but stable/ocata branch still exists for them. Could you please delete these? [3]
2. On the Release Team meeting [1] we were hinted that with the newer version of gerrit (that was installed at the end of November) some automation is possible through gerrit API in the future. Can I get some help about where should I start with the automation? Which repository should I look, where can the deletion being triggered ("similarly like branch creation")?
Thanks in advance,
Előd
[1]
http://eavesdrop.openstack.org/meetings/releaseteam/2020/releaseteam.2020-10... [2]
https://opendev.org/openstack/releases/src/commit/eb381492da3f7c826c35b9f147... [3] http://paste.openstack.org/show/801992/
-- Hervé Beraud Senior Software Engineer at Red Hat irc: hberaud https://github.com/4383/ https://twitter.com/4383hberaud -----BEGIN PGP SIGNATURE----- wsFcBAABCAAQBQJb4AwCCRAHwXRBNkGNegAALSkQAHrotwCiL3VMwDR0vcja10Q+ Kf31yCutl5bAlS7tOKpPQ9XN4oC0ZSThyNNFVrg8ail0SczHXsC4rOrsPblgGRN+ RQLoCm2eO1AkB0ubCYLaq0XqSaO+Uk81QxAPkyPCEGT6SRxXr2lhADK0T86kBnMP F8RvGolu3EFjlqCVgeOZaR51PqwUlEhZXZuuNKrWZXg/oRiY4811GmnvzmUhgK5G 5+f8mUg74hfjDbR2VhjTeaLKp0PhskjOIKY3vqHXofLuaqFDD+WrAy/NgDGvN22g glGfj472T3xyHnUzM8ILgAGSghfzZF5Skj2qEeci9cB6K3Hm3osj+PbvfsXE/7Kw m/xtm+FjnaywZEv54uCmVIzQsRIm1qJscu20Qw6Q0UiPpDFqD7O6tWSRKdX11UTZ hwVQTMh9AKQDBEh2W9nnFi9kzSSNu4OQ1dRMcYHWfd9BEkccezxHwUM4Xyov5Fe0 qnbfzTB1tYkjU78loMWFaLa00ftSxP/DtQ//iYVyfVNfcCwfDszXLOqlkvGmY1/Y F1ON0ONekDZkGJsDoS6QdiUSn8RZ2mHArGEWMV00EV5DCIbCXRvywXV43ckx8Z+3 B8qUJhBqJ8RS2F+vTs3DTaXqcktgJ4UkhYC2c1gImcPRyGrK9VY0sCT+1iA+wp/O v6rDpkeNksZ9fFSyoY2o =ECSj -----END PGP SIGNATURE-----
When Extended Maintenance process was created, there was no clear decision whether EOL (and branch deletion) is needed afterwards or not. But one thing we saw is that if a branch was EOL'd and still open, then patches arrived, and even some got approved which caused errors + confusions. Also, if the branch is open and periodic jobs are not deleted, then those are still running day by day. In this case only the branch deletion is the solution (as clearly those branches cannot accept job fixing patches). Though, you are right that if we don't even tag a branch with '$series-eol', then the above issue does not come. So theoretically we could forget about the 'eol' process, it would not cause any issue in my understanding. The periodic jobs needs to be deleted from .zuul.yaml of course, and maybe some other cleanup, otherwise it is possible. That's true. I can accept this, and this was my concept, too, in the beginning. Előd On 2021. 01. 27. 11:20, Herve Beraud wrote:
Le mer. 27 janv. 2021 à 10:55, Stephen Finucane <stephenfin@redhat.com <mailto:stephenfin@redhat.com>> a écrit :
On Tue, 2021-01-26 at 17:52 +0100, Előd Illés wrote: Hi Infra Team!
In October there was a discussion at Release Team meeting [1] about what can we do with the old, already EOL'd but not yet deleted branches (this is possible since with the Extended Maintenance process the general/"mass" EOL'ing was stopped and tagging a project branch EOL does not delete the branch anymore).
Not an answer but rather a question for my own understanding: what is the advantage of deleting branches? I understand that these things would no longer maintained and the gates will slowly break,
AFAIK this is mostly to avoid issues with gates/zuul.
but they're still relatively useful as a reference to explore project history and it's not like branches are expensive in git.
Tags can be used to dig in the related history.
Stephen
Related to this, I would like to ask two things:
1. I've used the list_eol_stale_branches.sh [2] script to get the list of such not-yet-deleted branches for Ocata [3]. They are all tagged with 'ocata-eol', but stable/ocata branch still exists for them. Could you please delete these? [3]
2. On the Release Team meeting [1] we were hinted that with the newer version of gerrit (that was installed at the end of November) some automation is possible through gerrit API in the future. Can I get some help about where should I start with the automation? Which repository should I look, where can the deletion being triggered ("similarly like branch creation")?
Thanks in advance,
Előd
[1] http://eavesdrop.openstack.org/meetings/releaseteam/2020/releaseteam.2020-10... [2] https://opendev.org/openstack/releases/src/commit/eb381492da3f7c826c35b9f147... [3] http://paste.openstack.org/show/801992/
-- Hervé Beraud Senior Software Engineer at Red Hat irc: hberaud https://github.com/4383/ https://twitter.com/4383hberaud -----BEGIN PGP SIGNATURE-----
wsFcBAABCAAQBQJb4AwCCRAHwXRBNkGNegAALSkQAHrotwCiL3VMwDR0vcja10Q+ Kf31yCutl5bAlS7tOKpPQ9XN4oC0ZSThyNNFVrg8ail0SczHXsC4rOrsPblgGRN+ RQLoCm2eO1AkB0ubCYLaq0XqSaO+Uk81QxAPkyPCEGT6SRxXr2lhADK0T86kBnMP F8RvGolu3EFjlqCVgeOZaR51PqwUlEhZXZuuNKrWZXg/oRiY4811GmnvzmUhgK5G 5+f8mUg74hfjDbR2VhjTeaLKp0PhskjOIKY3vqHXofLuaqFDD+WrAy/NgDGvN22g glGfj472T3xyHnUzM8ILgAGSghfzZF5Skj2qEeci9cB6K3Hm3osj+PbvfsXE/7Kw m/xtm+FjnaywZEv54uCmVIzQsRIm1qJscu20Qw6Q0UiPpDFqD7O6tWSRKdX11UTZ hwVQTMh9AKQDBEh2W9nnFi9kzSSNu4OQ1dRMcYHWfd9BEkccezxHwUM4Xyov5Fe0 qnbfzTB1tYkjU78loMWFaLa00ftSxP/DtQ//iYVyfVNfcCwfDszXLOqlkvGmY1/Y F1ON0ONekDZkGJsDoS6QdiUSn8RZ2mHArGEWMV00EV5DCIbCXRvywXV43ckx8Z+3 B8qUJhBqJ8RS2F+vTs3DTaXqcktgJ4UkhYC2c1gImcPRyGrK9VY0sCT+1iA+wp/O v6rDpkeNksZ9fFSyoY2o =ECSj -----END PGP SIGNATURE-----
On 2021-01-27 12:06:41 +0100 (+0100), Előd Illés wrote:
When Extended Maintenance process was created, there was no clear decision whether EOL (and branch deletion) is needed afterwards or not. But one thing we saw is that if a branch was EOL'd and still open, then patches arrived, and even some got approved which caused errors + confusions. Also, if the branch is open and periodic jobs are not deleted, then those are still running day by day. In this case only the branch deletion is the solution (as clearly those branches cannot accept job fixing patches).
Though, you are right that if we don't even tag a branch with '$series-eol', then the above issue does not come. So theoretically we could forget about the 'eol' process, it would not cause any issue in my understanding. The periodic jobs needs to be deleted from .zuul.yaml of course, and maybe some other cleanup, otherwise it is possible. That's true. I can accept this, and this was my concept, too, in the beginning. [...]
Deleting the branch serves a couple of purposes for visibility: first, as noted, people will get an error when trying to push a change for them; but second, consumers of those branches will actually no longer be able to pull from them nor check them out, and will have to switch to checking out the -eol tag for them instead, which makes it very clear that we are no longer developing on them. -- Jeremy Stanley
participants (4)
-
Előd Illés
-
Herve Beraud
-
Jeremy Stanley
-
Stephen Finucane