Issue #6519
closedpulp_rpm RpmCopy api - copy method fails to copy all rpm packages with dependency solving 'on'
Added by jdjeffers over 4 years ago. Updated over 4 years ago.
Description
While using the copy api with a yum repo using https://jlsherrill.fedorapeople.org/fake-repos/needed-errata/ as the upstream URL, only 31 of 32 rpm packages were copied into the target repository.
There are 2 dummy packages by the name of 'walrus'. The 'downlevel' rpm of 'walrus' was not copied.
Updated by daviddavis over 4 years ago
- Project changed from Pulp to RPM Support
- Subject changed from pulp_rpm RpmCopy api - copy method fails to copy all rpm packages with depeendency solving 'on' to pulp_rpm RpmCopy api - copy method fails to copy all rpm packages with dependency solving 'on'
Updated by jdjeffers over 4 years ago
Addition information - this is the data JSON provided to the copy api:
{
"config": [
{
"source_repo_version": "/pulp/api/v3/repositories/rpm/rpm/bbe8309a-4dff-416a-870e-03b2ecd63143/versions/1/",
"dest_repo": "/pulp/api/v3/repositories/rpm/rpm/514763db-7be8-4e7a-88df-b55daccafd51/",
"content": [
"/pulp/api/v3/content/rpm/packages/129908cf-8822-45e1-90f9-e50a0f1f4edc/",
"/pulp/api/v3/content/rpm/packages/1bbeaf26-45ac-4972-822c-8d3d13f0f93e/",
"/pulp/api/v3/content/rpm/packages/1d75f794-e9ef-454a-aae7-f7088bb35cb6/",
"/pulp/api/v3/content/rpm/packages/1f1edd14-0c88-4617-9a1c-5bc5ed4a4d43/",
"/pulp/api/v3/content/rpm/packages/27ed7731-3e6a-451c-8d9f-082b29992a3d/",
"/pulp/api/v3/content/rpm/packages/37f8a1da-a798-4ef1-9647-7320585a2f3b/",
"/pulp/api/v3/content/rpm/packages/41ea27eb-f12b-40f0-9b2a-f22a097e83c1/",
"/pulp/api/v3/content/rpm/packages/4829aabb-c568-443b-8224-0addd385eb8b/",
"/pulp/api/v3/content/rpm/packages/5d1ee916-65e8-40b5-a427-20737470dd45/",
"/pulp/api/v3/content/rpm/packages/616ac763-8110-44de-9906-bc2e5f8436f0/",
"/pulp/api/v3/content/rpm/packages/6e54e885-715a-4355-904b-0a39fb8c0673/",
"/pulp/api/v3/content/rpm/packages/783acf38-6858-415b-b204-ff3251645ae4/",
"/pulp/api/v3/content/rpm/packages/78f1a2a7-67e5-4af1-ac28-110f592cfdaf/",
"/pulp/api/v3/content/rpm/packages/7ba5ae27-63ae-48c4-8102-0dc62c4ba423/",
"/pulp/api/v3/content/rpm/packages/8465db44-8c9b-4491-b78b-087d17da9abb/",
"/pulp/api/v3/content/rpm/packages/90a7b5a1-1730-46b9-83c3-a94b9e11c993/",
"/pulp/api/v3/content/rpm/packages/9612b028-ec14-4ec8-a2b0-1fd2ad0bde74/",
"/pulp/api/v3/content/rpm/packages/9a70c1ce-f0f9-4d29-914e-fbd4318fd38a/",
"/pulp/api/v3/content/rpm/packages/a07b18f9-384b-4dd7-99d5-db0715fc0d8d/",
"/pulp/api/v3/content/rpm/packages/a15fdb2a-9f15-4411-9961-7cb14d0023ef/",
"/pulp/api/v3/content/rpm/packages/b30f35f9-d8d1-40a2-94aa-eeb10d4574b2/",
"/pulp/api/v3/content/rpm/packages/bc8d47a6-d5b5-4173-b7e8-f680b7dfddcf/",
"/pulp/api/v3/content/rpm/packages/c096bfa5-e0e0-451b-9e9e-476283734391/",
"/pulp/api/v3/content/rpm/packages/c83856a0-c726-470b-825e-065c80e19b7d/",
"/pulp/api/v3/content/rpm/packages/cc7b6aa2-01d1-4400-ba4b-3e41f07f47c6/",
"/pulp/api/v3/content/rpm/packages/cf18314e-ddd5-44a7-a35a-c53144913997/",
"/pulp/api/v3/content/rpm/packages/d9a11d6b-cd8c-464c-93dd-928a88d3989f/",
"/pulp/api/v3/content/rpm/packages/da391b20-0940-49ce-b207-33f823349374/",
"/pulp/api/v3/content/rpm/packages/dfa4daef-cb76-4c49-a325-b14ff4a5c42b/",
"/pulp/api/v3/content/rpm/packages/e33690bb-e85b-46b6-b8ca-bc26cf474013/",
"/pulp/api/v3/content/rpm/packages/e632229b-73b7-4317-a91c-3d2e9dfca632/",
"/pulp/api/v3/content/rpm/packages/f734fd87-89af-4475-b2ed-00cd064a67bd/",
"/pulp/api/v3/content/rpm/advisories/c4fdc73a-0260-40b1-880c-26b138b41f6f/",
"/pulp/api/v3/content/rpm/advisories/7f1d0bbf-461f-41a9-a8f8-fb4f2fc3c424/",
"/pulp/api/v3/content/rpm/advisories/abad23a4-5fda-4bdc-a632-5616909836ce/",
"/pulp/api/v3/content/rpm/advisories/db1ccbed-2ff3-4c76-98ea-9d759a2b8d71/"
]
}
],
"dependency_solving": true
}
Updated by ttereshc over 4 years ago
- Triaged changed from No to Yes
- Sprint set to Sprint 71
Updated by ggainey over 4 years ago
- Status changed from NEW to ASSIGNED
- Assignee set to ggainey
From IRC discussion:
<ttereshc> dalley, ggainey, we have a dep solving issue https://pulp.plan.io/issues/6519 which needs to be fixed in the upcoming sprint (preferably in the first half, to leave some time for release). IIRC, dalley is still the only knowledge carrier for the dep solving and maybe ggainey investigated/fixed something in pulp2 related to dep solving. So I think we are in need to find someone apart from dalley who can solve dep solving issues and I think this one is
<ttereshc> a good candidate for that. I wonder if ggainey is interested or if someone else is willing to try it out. For the vacation enablement purposes (I think rchan calls it that way), we need to spread the knowledge :)
<ggainey> haven't we seen this before?
<ttereshc> it was filed couple of weeks ago
<ggainey> depsolv finds the "most appropriate" rpm and thus we fail to find the 'downlevel' one? it's ringing a faint bell
<ttereshc> define "before"
<ggainey> like, months ago
<ttereshc> maybe, was it pulp2 or pulp3
<ggainey> i'm cudgeling my branemeetz
<ggainey> pulp2
<ggainey> iirc
<ttereshc> ok, so this is pulp3
<ttereshc> so maybe it's a simialr problem
<ttereshc> ggainey, do I read it correctly that you are interested in solving that? :P
<ggainey> heh - sure
<ggainey> i'm already digging thru old notes to see if I can remember exactly what was happening in The Before Times
<dalley> ggainey, I can tell you where the problem likely is, or at least where to start looking
<ggainey> dalley: yeah - let's write things down in the issue
<ttereshc> ggainey, dalley, it's not urgent yet and can be done next week, I just don't want you to context switch immediately.
<ttereshc> ggainey, thanks for agreeing to work on it, dalley - thanks for not solving it yourself but spread the knowledge :)
<ggainey> heheh
<ggainey> dalley: could this be related to https://bugzilla.redhat.com/show_bug.cgi?id=1784165 ? (because I think that's the bug I'm thinking of)
<ggainey> from the related issue : "This works, until we get into the situation described here, where we have a unit (chimpanzee) that has a dependency (walrus) that we haven't seen yet, and which is in the source repo. Depsolving in isolation results in chimpanzee pulling in the most up-to-date version(s) of its dependencies that the Solver can find, resulting in walrus-5.21 being pulled in from the repo."
<ggainey> hrm
<dalley> ttereshc, ggainey I'm about 90% sure the issue can be fixed by just deleting this "else" :) https://github.com/pulp/pulp_rpm/blob/master/pulp_rpm/app/depsolving.py#L841
<ggainey> may not be the exact problem - but it's def the area/context
<ttereshc> you are both troubles :)
<ttereshc> I'm going back to the migraiton plugin work
<dalley> ggainey, to be clear, delete the else -- and always do passthrough[repo].add(unit.pk)
<ggainey> dalley: while I have No Idea if that's the answer, that code is *definitely* the place I was thinking of looking
<ggainey> ahhh, interesting
<ggainey> a'ight, drop that in the issue and I'll pick it up and start testing
<dalley> basically, all units that were passed in, should come out the other side (be passed through) -- we shouldn't make a special case for just the units that are not dependency-solved
<ggainey> yeah zacly
<ggainey> "special cases" just make for "special bugz" :)
Updated by pulpbot over 4 years ago
- Status changed from ASSIGNED to POST
Added by ggainey over 4 years ago
Updated by ggainey over 4 years ago
- Status changed from POST to MODIFIED
Applied in changeset d2f61b8468f2f1dd4a1ffb975fcc7d20026a7496.
Added by ggainey over 4 years ago
Revision d401dd3a | View on GitHub
Fixed copy to always include explicitly-requested packages.
Added a recursive-copy test to notice the problem.
fixes #6519
(cherry picked from commit d2f61b8468f2f1dd4a1ffb975fcc7d20026a7496)
Updated by ggainey over 4 years ago
Applied in changeset d401dd3a550fbda62bc44f088717d37a47403c4c.
Updated by ttereshc over 4 years ago
- Status changed from MODIFIED to CLOSED - CURRENTRELEASE
- Sprint/Milestone set to Pulp RPM 3.3.1
Fixed copy to always include explicitly-requested packages.
Added a recursive-copy test to notice the problem.
fixes #6519