Description of problem: [RHCS Containers] Setting environment variable TCMALLOC_MAX_TOTAL_THREAD_CACHE_BYTES for all the ceph daemons I have verified it is set properly in rpm-based installations maybe because of /etc/sysconfig/ceph has this variable set in rpm-based installation. Coming to containers we have to set the following options: 1. for osds: In all.yml ceph_tcmalloc_max_total_thread_cache: 134217728 2. For mons: In mons.yml ceph_mon_docker_extra_env: "-e TCMALLOC_MAX_TOTAL_THREAD_CACHE_BYTES={{ ceph_tcmalloc_max_total_thread_cache }}" 3. For mgrs: In mgrs.yml ceph_mgr_docker_extra_env: "-e TCMALLOC_MAX_TOTAL_THREAD_CACHE_BYTES={{ ceph_tcmalloc_max_total_thread_cache }}" 4. For rgws: In rgws.yml ceph_rgw_docker_extra_env: "-e TCMALLOC_MAX_TOTAL_THREAD_CACHE_BYTES={{ ceph_tcmalloc_max_total_thread_cache }}" We need a fix where this is all set by default in RHCS 4.x containers clusters and we do not have to set it manually. The default value for this variable will be 134217728 (128MB). For RHCS 5.0 we are handling this here - https://bugzilla.redhat.com/show_bug.cgi?id=1970597 Version-Release number of selected component (if applicable): RHCS 4.2
(In reply to Vikhyat Umrao from comment #0) > Description of problem: > > [RHCS Containers] Setting environment variable > TCMALLOC_MAX_TOTAL_THREAD_CACHE_BYTES for all the ceph daemons > > I have verified it is set properly in rpm-based installations maybe because > of /etc/sysconfig/ceph has this variable set in rpm-based installation. > > Coming to containers we have to set the following options: > > 1. for osds: In all.yml > ceph_tcmalloc_max_total_thread_cache: 134217728 > > 2. For mons: In mons.yml > ceph_mon_docker_extra_env: "-e TCMALLOC_MAX_TOTAL_THREAD_CACHE_BYTES={{ > ceph_tcmalloc_max_total_thread_cache }}" > > > 3. For mgrs: In mgrs.yml > ceph_mgr_docker_extra_env: "-e TCMALLOC_MAX_TOTAL_THREAD_CACHE_BYTES={{ > ceph_tcmalloc_max_total_thread_cache }}" > > > 4. For rgws: In rgws.yml > ceph_rgw_docker_extra_env: "-e TCMALLOC_MAX_TOTAL_THREAD_CACHE_BYTES={{ > ceph_tcmalloc_max_total_thread_cache }}" > > We need a fix where this is all set by default in RHCS 4.x containers > clusters and we do not have to set it manually. The default value for this > variable will be 134217728 (128MB). > > For RHCS 5.0 we are handling this here - > https://bugzilla.redhat.com/show_bug.cgi?id=1970597 > > Version-Release number of selected component (if applicable): > RHCS 4.2 I forgot to mention MDS we also need for MDS :)
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory (Red Hat Ceph Storage 4.2 Bug Fix update), and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHBA-2021:3670