NETOBSERV-1091: remove CO-RE file and extensions as that causes douple allocations #133
Conversation
Signed-off-by: msherif1234 <mmahmoud@redhat.com>
|
/ok-to-test |
openshift-ci
bot
added
the
ok-to-test
|
New image: quay.io/netobserv/netobserv-ebpf-agent:5fee00e. It will expire after two weeks. |
Codecov Report
@@ Coverage Diff @@
## main #133 +/- ##
=======================================
Coverage 40.60% 40.60%
=======================================
Files 31 31
Lines 2054 2054
=======================================
Hits 834 834
Misses 1181 1181
Partials 39 39
Flags with carried forward coverage won't be shown. Click here to find out more. |
msherif1234
changed the title
|
@msherif1234: This pull request references NETOBSERV-1091 which is a valid jira issue. In response to this:
Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository. |
|
@msherif1234: This pull request references NETOBSERV-1091 which is a valid jira issue. In response to this:
Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository. |
|
@msherif1234: This pull request references NETOBSERV-1091 which is a valid jira issue. In response to this:
Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository. |
| @@ -13,11 +13,11 @@ | |||
| until an entry is available. | |||
| 4) When hash collision is detected, we send the new entry to userpace via ringbuffer. | |||
| */ | |||
| #define BPF_NO_PRESERVE_ACCESS_INDEX | |||
There was a problem hiding this comment.
can you add a comment explaining briefly the reason (or just linking to the JIRA and/or libpbf issue) ?
|
I'm fine to merge it and also backport on 1.3 |
I linked the issue as well as the PR that trigger this in the PR description shouldn't that be enough ? |
|
/label qe-approved |
|
/lgtm |
|
/approve |
|
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: jotak The full list of commands accepted by this bot can be found here. The pull request process is described here
Needs approval from an approver in each of these files:
Approvers can indicate their approval by writing |
* Optimize ebpf agent map memory usage - switch to use pointer to metric instead of metric - manuall trigger GC after flow eviction complete Signed-off-by: msherif1234 <mmahmoud@redhat.com> * Fix memory and cpu scale issue work around in #133 following up on cilium/ebpf#1063 it seems we have a way to fix resources issues Signed-off-by: msherif1234 <mmahmoud@redhat.com> (cherry picked from commit b9c9a03) --------- Signed-off-by: msherif1234 <mmahmoud@redhat.com>
it was noticed cpu and memory are higher than 1.2 release
that change was triggered by #110
it seems using co-re results in a significant memory and cpu increase cilium/ebpf#1063 is opened against cilium/ebpf to track this, issue has been seen even with latest version of the lib, we don't use CO-RE apis at all so for now we will remove it and once we have fix we will re-enable
diag on left side with this PR changes the diag on right is with 1.3 baseline
