Note: This is a public test instance of Red Hat Bugzilla. The data contained within is a snapshot of the live data so any changes you make will not be reflected in the production Bugzilla. Email is disabled so feel free to test any aspect of the site that you want. File any problems you find or give feedback at bugzilla.redhat.com.
Bug 1591808
| Summary: | 389-ds-base bogusly provides libjemalloc.so.2()(64bit) | ||
|---|---|---|---|
| Product: | [Fedora] Fedora | Reporter: | Jun Aruga <jaruga> |
| Component: | 389-ds-base | Assignee: | mreynolds |
| Status: | CLOSED ERRATA | QA Contact: | Fedora Extras Quality Assurance <extras-qa> |
| Severity: | unspecified | Docs Contact: | |
| Priority: | unspecified | ||
| Version: | rawhide | CC: | fabian.deutsch, fpercoco, jal233, mhroncok, mreynolds, nathans, rmeggins, vashirov, vondruch |
| Target Milestone: | --- | ||
| Target Release: | --- | ||
| Hardware: | Unspecified | ||
| OS: | Unspecified | ||
| Whiteboard: | |||
| Fixed In Version: | 389-ds-base-1.4.0.11-1.fc29 389-ds-base-1.4.0.11-2.fc28 | Doc Type: | If docs needed, set a value |
| Doc Text: | Story Points: | --- | |
| Clone Of: | Environment: | ||
| Last Closed: | 2018-06-20 13:48:09 UTC | Type: | Bug |
| Regression: | --- | Mount Type: | --- |
| Documentation: | --- | CRM: | |
| Verified Versions: | Category: | --- | |
| oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |
| Cloudforms Team: | --- | Target Upstream Version: | |
| Embargoed: | |||
|
Description
Jun Aruga
2018-06-15 15:37:58 UTC
python-redis won't build from sources because of this. $ rpm -qp --requires redis-4.0.10-1.fc29.x86_64.rpm | grep jemalloc libjemalloc.so.2()(64bit) $ rpm -qp --requires redis-4.0.9-1.fc29.x86_64.rpm | grep jemalloc libjemalloc.so.2()(64bit) This is not a bug in redis, but in 389-ds-base: $ dnf repoquery --disablerepo='*' --enablerepo='rawhide' --whatprovides 'libjemalloc.so.2()(64bit)' 389-ds-base-0:1.4.0.10-2.fc29.x86_64 389-ds-base-libs-0:1.4.0.10-2.fc29.x86_64 jemalloc-0:5.1.0-1.fc29.x86_64 This should now be fixed in 389-ds-base-1.4.0.11-1 389-ds-base-1.4.0.11-1.fc28 has been submitted as an update to Fedora 28. https://bodhi.fedoraproject.org/updates/FEDORA-2018-b3f2f29e65 Thanks for the working! rubygem-redis-namespace on rawhide works now. (In reply to Jun Aruga from comment #5) > Thanks for the working! > rubygem-redis-namespace on rawhide works now. No problem, sorry for the inconvenience. 389-ds-base-1.4.0.11-1.fc28 has been pushed to the Fedora 28 testing repository. If problems still persist, please make note of it in this bug report. See https://fedoraproject.org/wiki/QA:Updates_Testing for instructions on how to install test updates. You can provide feedback for this update here: https://bodhi.fedoraproject.org/updates/FEDORA-2018-b3f2f29e65 @Mark I have a few notes. * Why is the bundled jemalloc used on the first place? * Even after the change, there is still jemalloc referenced in -devel and -libs subpackages. Is the bundled jemalloc dynamically linked? Why? * Even though the bundling is not strictly prohibited as it used to be, you still should be using "Provides: bundled(<libname>) = <version>" [1]. [1] https://fedoraproject.org/wiki/Packaging:Guidelines#Bundling_and_Duplication_of_system_libraries (In reply to Vít Ondruch from comment #8) > @Mark > > I have a few notes. > > * Why is the bundled jemalloc used on the first place? Because glibc memory allocator is awful in regards to the way we use memory in DS. We see massive RSS memory growth with glibc, but not with jemalloc (or tcmalloc). This has caused countless customer escalations. So we have no choice but to use jemalloc until the glibc allocator gets improved. There are bugs opened for this. > > * Even after the change, there is still jemalloc referenced in -devel and > -libs subpackages. Is the bundled jemalloc dynamically linked? Why? We are aware of this and are looking into it... > > * Even though the bundling is not strictly prohibited as it used to be, you > still should be using "Provides: bundled(<libname>) = <version>" [1]. We will look into this as well... > > [1] > https://fedoraproject.org/wiki/Packaging: > Guidelines#Bundling_and_Duplication_of_system_libraries Thanks for the pointer! (In reply to mreynolds from comment #9) > (In reply to Vít Ondruch from comment #8) > > * Why is the bundled jemalloc used on the first place? > > Because glibc memory allocator is awful in regards to the way we use memory > in DS. We see massive RSS memory growth with glibc, but not with jemalloc > (or tcmalloc). This has caused countless customer escalations. > > So we have no choice but to use jemalloc until the glibc allocator gets > improved. There are bugs opened for this. My question was more about the "bundling". 389-ds-base-1.4.0.11-2.fc28 has been submitted as an update to Fedora 28. https://bodhi.fedoraproject.org/updates/FEDORA-2018-a00a3e9b36 389-ds-base-1.4.0.11-2.fc28 has been pushed to the Fedora 28 testing repository. If problems still persist, please make note of it in this bug report. See https://fedoraproject.org/wiki/QA:Updates_Testing for instructions on how to install test updates. You can provide feedback for this update here: https://bodhi.fedoraproject.org/updates/FEDORA-2018-a00a3e9b36 389-ds-base-1.4.0.11-2.fc28 has been pushed to the Fedora 28 stable repository. If problems still persist, please make note of it in this bug report. |