Hi Mercury, That is indeed the reason. Bitmaps as opposed to vectors/pools rely on inline functions so if they grow the api consumer needs to update its pointer to the new underlying data structure.
As for how to debug these, unfortunately I don’t have a good recipe. In this particular case, your observation that it works with 1 process instead of 2 (master off vs on) made me think the issue is related to the vls logic that deals with shared sessions. It was easy to confirm that in the latter case accepted sessions uses more memory with clib_mem_get_heap_usage(). But it took me quite some time to find the actual spot as I focused on the accepted sessions at first instead of the listener. Hope this helps! Regards, Florin > On Dec 9, 2021, at 1:35 AM, mercury noah <mercury124...@gmail.com> wrote: > > Hi Florin, > > We tried the patch and it works for us as the jira ticket metioned, thanks a > lot, > Sorry for replying so late, cuz I spent some time to find the > reason(clib_bitmap_set().clib_bitmap_vec_validate()), > It would be more helpful if you would like to share some ideas about how to > find it at your convenience, > > Thanks, > Mercury > > >
-=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply Online (#20612): https://lists.fd.io/g/vpp-dev/message/20612 Mute This Topic: https://lists.fd.io/mt/87398061/21656 Mute #vpp-hoststack:https://lists.fd.io/g/vpp-dev/mutehashtag/vpp-hoststack Mute #hoststack:https://lists.fd.io/g/vpp-dev/mutehashtag/hoststack Mute #vppcom:https://lists.fd.io/g/vpp-dev/mutehashtag/vppcom Mute #nginx:https://lists.fd.io/g/vpp-dev/mutehashtag/nginx Group Owner: vpp-dev+ow...@lists.fd.io Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub [arch...@mail-archive.com] -=-=-=-=-=-=-=-=-=-=-=-
