RHBA-2017:3387 - Bug Fix Advisory

Topic

Red Hat Ceph Storage 3.0 is now available.

Description

Red Hat Ceph Storage is a scalable, open, software-defined storage platform that combines the most stable version of the Ceph storage system with a Ceph management platform, deployment utilities, and support services.

Bug Fixes and Enhancements:

For detailed information on changes in this release, see the Red Hat Ceph
Storage 3.0 Release Notes available at:

https://access.redhat.com/documentation/en-us/red_hat_ceph_storage/3.0/html-single/release_notes/

Solution

Before applying this update, make sure all previously released errata
relevant to your system have been applied.

For details on how to apply this update, refer to:

https://access.redhat.com/articles/11258

Affected Products

• Red Hat Enterprise Linux Server 7 x86_64
• Red Hat Ceph Storage MON 3 x86_64
• Red Hat Ceph Storage OSD 3 x86_64

Fixes

• BZ - 1267636 - [RFE] Add config option to cause the mons to complain to the cluster log when any client older than the configured version tries to authenticate
• BZ - 1300803 - [1.3.1Async - 0.94.3-6] qemu xfs tests - osd/ReplicatedPG.cc: 2351: FAILED assert(0 == "out of order op")
• BZ - 1305670 - RFE: add --pg-num to osdmaptool --test-map-pgs to allow the user to experiment with other pg counts
• BZ - 1330732 - [RFE] Adding an optional limit to RDB snapshot
• BZ - 1332083 - log [ERR] : OSD full dropping all updates 99% full followed by FAILED assert(0 == "unexpected error") : ENOSPC handling not implemented
• BZ - 1337018 - [RFE] filestore: randomize split threshold
• BZ - 1343189 - remove dependency on libfcgi in radosgw and librgw2 packages
• BZ - 1344212 - Unable to Read the Image Metadata from the Slave Node
• BZ - 1344262 - Disabling a feature from Master Node, is not getting reflected on Slave Node
• BZ - 1353877 - Try to Disable image feature from Slave Node is resulting in indefinite wait
• BZ - 1362014 - 2.0: rbd bench-write: segmentation fault when value of "--io-size" is greater than or equal to image size
• BZ - 1362055 - 2.0: "rbd du <non existing image>" does not print an error message
• BZ - 1365034 - [rbd-mirror] - failed to commit journal event
• BZ - 1366807 - [RFE] ceph-ansible: remove MON and OSD nodes
• BZ - 1369586 - ceph osd pool create uses 'rule_id' vs 'ruleset' when ruleset name specified during pool creation
• BZ - 1375078 - [ubuntu] Add Repositories with Debug packages
• BZ - 1377573 - ceph-objectstore-tool crashing in basic operation
• BZ - 1387437 - Object Gateway does not support absolute URL
• BZ - 1388541 - [rbd-mirror] : snapshot name is different than given name( '.rbd-mirror.0f454219-dd15-49ba-968a-f822833d46e8.87ea3ebe-1c4a-4bdb-8942-245196739d51' instead of 'snap<number>')
• BZ - 1388703 - Restart or reload of the rbd-target-gw service fails during alua configuration
• BZ - 1406523 - [RFE] RGW Metadata Log Trimming for MultisiteV2
• BZ - 1410923 - rbd export/export-diff commands not honoring '--rbd-concurrent-management-ops' option, option appears to have no effect on performance
• BZ - 1415557 - Do not alert if cephx is not set in ceph.conf
• BZ - 1415778 - [RFE] Support for dm-cache
• BZ - 1417163 - RFE: Support for bucket owner to grant cross account bucket permissions
• BZ - 1427512 - [RFE] add option to be enable mount Cephfs which reflects used pool characteristics, not ceph df global
• BZ - 1428634 - [RFE] [rbd CLI]: ability to mass promote / demote all mirrored images within a pool
• BZ - 1434175 - [RFE] automate setting up online repositories for ceph deployments on debian nodes
• BZ - 1436782 - [RFE] package ceph-ansible for Ubuntu
• BZ - 1441693 - [RFE] Addition of online osd 'omap'compaction command
• BZ - 1446449 - tcmu-runner: reports remote ports incorrectly which leads to windows/esx failing to failover
• BZ - 1449767 - [rbd-mirror] : mirror status is being checked by mirror service for new pools created
• BZ - 1450754 - RHCS ceph-ansible rolling upgrade sets/unsets cluster flags between every OSD upgrade
• BZ - 1451936 - [RFE]: osd: deletes are performed inline during pg log processing
• BZ - 1452762 - [RFE] RGW implement ansible installation and setup of RGW NFS
• BZ - 1455711 - [RFE] OSD: Add heartbeat message for Jumbo Frames(MTU 9000)
• BZ - 1456036 - [RHCeph 2.3]: default osd crush tunables is set to firefly
• BZ - 1458224 - Improve check PGS when OSD daemons are restarted
• BZ - 1461041 - only log metadata on metadata master zone
• BZ - 1461551 - remove, disable, or warn about untested/undocumented playbooks
• BZ - 1462012 - [RFE] [RHCS 3.0] Change default filestore 'omap' backend to 'rocksdb' from 'leveldb'
• BZ - 1464955 - [RFE] Simplified CephFS client key creation
• BZ - 1464957 - [RFE] Simplified RBD client key creation
• BZ - 1464959 - [RFE]: Improved prioritization of OSD client I/O over recovery/backfill
• BZ - 1464964 - [RFE]: Avoid raising spurious health warnings during initial setup of cluster
• BZ - 1464966 - [RFE]: Ceph cluster should not create any pools by default
• BZ - 1464973 - [RFE]: Configuration option metadata
• BZ - 1464976 - [RFE] Revise CephFS cache configuration to operate in MB instead of inode count
• BZ - 1464980 - [RFE]: Improved/sanitized cluster log
• BZ - 1464981 - [RFE]: Sanitized health check format
• BZ - 1466054 - bump epoch in RHCEPH 3.0 layered product
• BZ - 1467352 - [RFE] Enable an easy method to delete objects in a Ceph pool if an OSD hit full_ratio
• BZ - 1467976 - ceph logs shouldn't display dmcrypt key
• BZ - 1469426 - Support clients configuration with external Ceph clusters
• BZ - 1469781 - Deploying multiple RGWs on different hosts mix up service creation
• BZ - 1470826 - RFE: NFS V3 support for RGW NFS
• BZ - 1470829 - [RFE]: compression of data at rest for RGW
• BZ - 1470837 - RFE: Bucket Policy compatibility with S3
• BZ - 1471035 - It is not possible to set MDS caps for the openstack_keys
• BZ - 1471152 - TASK [ceph-client : copy ceph admin keyring] failing when docker: true
• BZ - 1471241 - ship ceph-ansible and Ansible in RHCEPH 3
• BZ - 1471989 - Keys created by "create openstack key(s)" task are not readable by openstack users nova, cinder, etc.
• BZ - 1472464 - rgw multisite: objects encrypted with SSE-C fail to sync
• BZ - 1472465 - rgw multisite: objects encrypted with SSE-KMS do not sync correctly
• BZ - 1472874 - Stale bucket index entries are left over after object deletions due to invalid timespan issues
• BZ - 1473386 - rebase nfs-ganesha to 2.5
• BZ - 1473399 - [RFE]: rgw: nfs-ganesha store v4 clientids in RADOS
• BZ - 1473400 - [RFE]: rgw: nfs-ganesha store exports in RADOS
• BZ - 1473405 - Truncated S3 objects
• BZ - 1473410 - rgw: Part's index is not removed when versioning enabled
• BZ - 1474436 - [RFE] Radosgw-admin utility to list users
• BZ - 1475016 - RFE: S3 object expiration support
• BZ - 1475032 - RFE: Object granular encryption support
• BZ - 1475820 - allow multi dedicated journals for container deployment
• BZ - 1475982 - Package cephmetrics
• BZ - 1476210 - [ceph-ansible]: Bluestore is configured by default on the OSD's
• BZ - 1477300 - [RFE] dynamic bucket sharding
• BZ - 1477311 - Restrict --inconsistent-index flag from radosgw-admin bucket rm command because it can cause corrupted bucket index
• BZ - 1478071 - Set ceph_rhcs: true as default value in all.yml.sample
• BZ - 1478599 - older versions of lvm do not support --reportformat=json
• BZ - 1479138 - CephFS: fusermount utility not available to unmount fuse mount.
• BZ - 1479673 - [rbd-mirror] - Unable to write data on the promoted image from secondary rbd host
• BZ - 1479797 - do not ignore non-zero exit status when activating
• BZ - 1480182 - [CephFS] Posix test failed with EPERM error on Ceph-Fuse mount.
• BZ - 1480305 - ceph: Error: error connecting to the cluster: errno ENOTSUP in nova-compute.log
• BZ - 1481755 - rebase/update RBD iSCSI packages for RHCS 3.0
• BZ - 1481828 - [RFE] osd: osd_scrub_during_recovery only considers primary, not replicas
• BZ - 1482061 - [ceph-ansible] [ceph-container] : restart osds daemon task should be skipped for non osd node during cluster deployment
• BZ - 1482067 - [support] "ansible-playbook site.yml --limit <osds|clients|rgws>" will not create proper content of /etc/ceph/ceph.conf
• BZ - 1482462 - Set mds_address_block and radosgw_address_block like monitor_address_block
• BZ - 1482579 - introduce ceph-medic v1
• BZ - 1482700 - man pages needed for ceph-volume and ceph-volume-systemd
• BZ - 1482822 - [ceph-ansible] [ceph-container] : admin keyring must be copied to rgw node if copy_admin_key is set to true
• BZ - 1482897 - The switch-to-containers playbook fails when called from TripleO
• BZ - 1484002 - lvm_volumes dict entries must be unique logical volume names
• BZ - 1484083 - [ceph-iscsi-ansible]: ceph-iscsi-ansible can only be used with RHEL 7.3
• BZ - 1484189 - rhosp-director: HA overcloud with ceph and IPV6 fails
• BZ - 1485011 - cannot consistently use non-lv devices as journals
• BZ - 1485783 - [CephFS] Standby-Replay daemon is hanging in "resolve" state while trying to take over rank
• BZ - 1485795 - After installing ceph-medic on Ubuntu, got ImportError, for module tambo
• BZ - 1485887 - [UPDATES][OpenStack] Systemd ceph* units conflict with ceph* containers
• BZ - 1486062 - add support for condition like ceph_release_num
• BZ - 1486112 - config parameters for rados url provider not recognized
• BZ - 1486515 - ceph --version shows "rc"
• BZ - 1486534 - RFE: Ship / Enable ceph-mgr
• BZ - 1486535 - ship / enable ceph-mgr rest module
• BZ - 1488479 - RHCS upgrade 2.4 - 3.0: PGs are stuck at recovering and degraded state
• BZ - 1488519 - While renaming a non-existing rbd image returned segmentation fault
• BZ - 1488813 - [ceph-ansible] [ceph-container] : restart ceph mds/rgws task are failing on respective nodes before configuring mdss/rgws
• BZ - 1488966 - [iSCSI]: TCMU LIO checks for all supported rbd features on an image, when only exclusive-lock is required
• BZ - 1488999 - [ceph-ansible] [ceph-container] : admin keyring getting copied to mds node even if copy_admin_key is set to false
• BZ - 1489064 - prevent creation (of pools with) too many PGs
• BZ - 1489099 - OSP11 -> OSP12 upgrade: switch-from-non-containerized-to-containerized-ceph-daemons.yml fails with: "'list object' has no attribute 'stdout_lines'"
• BZ - 1489353 - OSP11 -> OSP12 upgrade: unable to rerun a failed major-upgrade-composable-steps-docker with ceph-ansible enabled because 'collect running osds' task fails
• BZ - 1489391 - RGW User creation inconsistency when given the same 'access_key' name
• BZ - 1489444 - No user friendly error messages from ceph-medic
• BZ - 1489447 - invalid --ssh-config path is not reported correctly, allowing checks
• BZ - 1489452 - [RGW:NFS]: Forceful deletion of directories fails for the first time with "directory not empty error" and passes the second time
• BZ - 1489461 - [ceph-mds] : mds process crashed during rm -rf from fuse client
• BZ - 1489690 - [ceph-container] - ceph-mgr - container getting exited saying /usr/bin/ceph-mgr: No such file or directory
• BZ - 1490188 - properly describe what release_version var should be set to during upgrade
• BZ - 1490278 - s3:GetBucketVersioning/PutBucketVersioning fails with 403
• BZ - 1490296 - [iSCSI]; rbd-target-api crashed on one of the GWs during disk creation
• BZ - 1490355 - [ceph-ansible] - shrink-osd failing when more than one osd is to be removed
• BZ - 1490814 - [Ceph-Fuse]: fuse service stopped and mount become unavailable during client IO's.
• BZ - 1491550 - [ceph-iscsi-cli]: gwcli fails to set auth without the "chap=" option
• BZ - 1491634 - caught signal (Illegal instruction) when activation bluestore OSD
• BZ - 1491682 - [ceph-iscsi-config]: New tpg creation should be blocked
• BZ - 1491844 - mgr keyring cannot be found
• BZ - 1491976 - [ceph-ansible] - Leftover entry of temporary rgw socket file path in conf file
• BZ - 1492032 - [ceph-iscsi-config]: rbd-target-gw fails to start due to missing 'config' object
• BZ - 1492056 - no package matching 'nfs-ganesha-rgw'
• BZ - 1492115 - [RFE] ceph-ansible: Shrinking containerized MON and OSD nodes
• BZ - 1492160 - [ceph-ansible]: mandatory variable check failed even though its enabled
• BZ - 1492165 - [ceph-ansible]: "take-over-existing-cluster.yml" fails as all.yml is not present anymore
• BZ - 1492223 - When yum.repos.d .repo file points to download-node-02.eng.bos.redhat.com, install fails.
• BZ - 1492224 - 'statleftover' is undefined
• BZ - 1492582 - [RGW:NFS]: Overwrite to an existing file fails with NFS version 3
• BZ - 1492675 - [ceph-ansible] - value of fetch_directory in rhcs.yml getting ignored
• BZ - 1492785 - rbd mirror pool promote <pool> hangs indefinitely
• BZ - 1492865 - [RHCeph 3.0/ 12.2.0-2.el7cp] test_client_pin (tasks.cephfs.test_client_limits.TestClientLimits)
• BZ - 1493231 - [ceph-ansible] - cluster initialization failing searching for (debian/redhat)_rhcs_repository.yml
• BZ - 1493552 - [ceph-ansible] - conf file generation failing on rgw nodes when nodes have different interface names
• BZ - 1493572 - Missing cephmetrics-ansible package
• BZ - 1493636 - Check in memory Ceph daemon version and compare with installed daemon version if different please report *Error*
• BZ - 1493750 - [RHCeph 3.0 / 12.2.0-2.el7cp] avc: denied { getattr } for pid=23270 comm="ceph-osd" path="/sys/dev/block/8:17" dev="sysfs"
• BZ - 1493896 - s3:GetBucketWebsite/PutBucketWebsite fails with 403
• BZ - 1493920 - [ceph-ansible] [ceph-container] : osd restart handler failing waiting PGs to be active+clean
• BZ - 1493934 - s3:GetBucketLocation bucket policy fails with 403
• BZ - 1493977 - rbd image-meta set <image-spec> conf_rbd_mirroring_replay_delay <delay in seconds>, delay is not taking effect.
• BZ - 1493985 - [ceph-ansible]: Failure in task "ceph-iscsi-gw : lock ssl file access to root only"
• BZ - 1494039 - OSD containers don't start after reboot
• BZ - 1494067 - cephmetrics-ansible 'Set collectd_dir' task
• BZ - 1494140 - s3:GetBucketCORS/s3:PutBucketCORS policy fails with 403
• BZ - 1494450 - [ceph-iscsi-cli]: gwcli should fail cleanly if new GW config does not match existing GW
• BZ - 1494543 - ceph-disk prepare cannot find partition that does exist when run inside containers
• BZ - 1494643 - [RHCeph 3.0/3.0.0-0.1.rc10.el7cp] Dont run include ceph_keys.yml when monitor's not in quorum
• BZ - 1494980 - pybind/ceph_volume_client: snapshot dir name hardcoded
• BZ - 1495110 - [ceph-iscsi-cli]: If cluster info is not available gwcli should fail gracefully
• BZ - 1495193 - cephmetrics SELinux policy
• BZ - 1495356 - nfs-ganesha service fails to start with systemd due to permission denied on /var/log/ganesha
• BZ - 1495521 - Rbd-mirror: Re-sync request against the "master" cluster failed to delete the image and sync the image from the "slave" cluster.
• BZ - 1495533 - [ceph-iscsi-cli]: JSON error seen when trying to delete a host entry
• BZ - 1496198 - 'Ceph - At A Glance - Capacity panel' missing units
• BZ - 1496460 - [ rgw ]: Uploads with SSE-KMS encryption do not include the encryption headers in response
• BZ - 1496519 - [iSCSI]: GW node reboot deletes tpg portal on the local node
• BZ - 1496585 - rgw_file: fix a special overlapped i/o case, increasing coverage
• BZ - 1496674 - [Perf] : 4k Random reads are off target by ~23%.
• BZ - 1496692 - [ceph-ansible] - value set for 'ceph_repository_type' as iso is getting ignored
• BZ - 1497194 - cephmetrics-ansible CDN repositories for ceph-grafana
• BZ - 1497198 - cephmetrics-ansible ceph-grafana PyYAML
• BZ - 1497332 - Free-space isn't released when deleting data-pool snapshots
• BZ - 1497853 - Failed CompleteMultipartUpload request does not release lock
• BZ - 1497895 - [ceph-ansible] [ceph-container] : filesystems pool generation failing - osd_pool_default_pg_num is undefined
• BZ - 1497959 - RHCS upgrade 2.4 - 3.0: Ceph-mgr daemons are not active after upgrade.
• BZ - 1497965 - RHCS upgrade 2.4 -3.0: HEALTH_WARN - all OSDs are running luminous or later but require_osd_release < luminous
• BZ - 1498180 - [Ceph-Ansible]:- Addition of nodes to the cluster fails in the build 3.0.0~rc14-2redhat1
• BZ - 1498218 - ceph-ansible RGW role restarts all RGWs simutaneously
• BZ - 1498474 - Don't reshard bucket in a multisite enviroment
• BZ - 1498490 - [ceph-ansible]: configuring iscsi gateways through ansible failed with "gateway_ip_list" error.
• BZ - 1498504 - 'Ceph OSD Information' - bad number in 'OSDs UP' panel
• BZ - 1498505 - [ceph-iscsi-config]: High number of disks in the config causes failure in loading the config
• BZ - 1498583 - [ceph-ansible] : admin, mon, mgr keyrings being copied to all nodes of the cluster
• BZ - 1498611 - [RHCeph 3.0/ 12.2.1-9.el7cp] src/common/interval_map.h: 161: FAILED assert(len > 0)
• BZ - 1498668 - RGWs quota stats simple fixes, esp. for multiple rgws.
• BZ - 1498825 - [ceph-iscsi-cli]: iSCSI GW creation failed with "INTERNAL SERVER ERROR"
• BZ - 1499132 - [ceph-ansible] RHCS 3.0 beta - Ubuntu iso installation fails
• BZ - 1499404 - [ceph-ansible] [ceph-container] : osd disk preparation failing for non-collocated scenario - 'None' has no attribute 'get'
• BZ - 1499412 - [ceph-ansible] - Duplication of rgw client section in conf file when rgw and nfs are collocated
• BZ - 1499734 - 'Top 5' panels on 'Ceph Pools' dashboard are not working properly
• BZ - 1499871 - [Ceph-ansible]: purge-cluster fails, is not zapping the disks
• BZ - 1500168 - [ rgw ]: S3 object encryption in Ubuntu multisite setup is not working.
• BZ - 1500281 - [ceph-ansible]: purging iSCSI GWs using ansible failed.
• BZ - 1500400 - [ceph-ansible] - 3.0.0-0.1.rc19 - NFS Package installation failed for RHEL ISO.
• BZ - 1500470 - 3.0 Upgrade: Container upgrade from 2.4 to 3.0 Failed with client.admin authentication error (1) Operation not permitted
• BZ - 1500669 - nfs-ganesha: FSAL_CEPH appears to allow conflicting locks
• BZ - 1500727 - [ceph-ansible] [ceph-container] : dmcrypt OSDs - open_encrypted_part failing
• BZ - 1500874 - [CephFS]:- All MDS crashing during subtree pinning ,IOs in progress and failover in Active-Active with standby setup
• BZ - 1501374 - Possible infinite loop in journal:get_tag_list class method will result in OSD suicide timeout
• BZ - 1501389 - [RGW-Compression] - Value passed to compression type is not validated.
• BZ - 1501408 - rgw multisite: multipart uploads fail to sync
• BZ - 1502178 - MDS cache trims entire un-pinned cache under read heavy workloads
• BZ - 1502763 - [RFE] rocksdb default options for filestore should include multithreaded compaction
• BZ - 1503251 - handler fail to restart containerized osd on collocation
• BZ - 1504166 - rbd-mirror doesn't restart after reboot
• BZ - 1504171 - mgr command fixes during upgrade
• BZ - 1504172 - compat with kraken-era clients (including kernels)
• BZ - 1504179 - several bluestore fixes
• BZ - 1504965 - [ubuntu] update Ansible to 2.4.1.0
• BZ - 1505011 - [Ceph-Ansible 3.0.3-1.el7cp] dont default restart machines during purge-cluster
• BZ - 1505048 - [ceph-ansible] [ceph-container] : upgrade of nfs failed when rolling update was tried only for nfs
• BZ - 1505365 - compat osdmap encoding does not reencode crush map
• BZ - 1505448 - nfs-ganesha: Bind_Addr parameter ignored
• BZ - 1505460 - librados C++ ABI compat break with jewel
• BZ - 1505485 - [RGW-Compression] - Data synced to secondary site is not compressed.
• BZ - 1505496 - [ceph-container] : containerized rbd mirroring daemon - unexpected peer client authentication error
• BZ - 1505502 - TCMALLOC_MAX_TOTAL_THREAD_CACHE_BYTES environment variable not passed to OSDs in case of Containerized Ceph
• BZ - 1505504 - space may be unreclaimable after inter-bucket copy (space leak)
• BZ - 1505512 - Enable RADOS based features in downstream Ganesha builds
• BZ - 1506132 - 'Ceph RGW Workload' editable dashboard
• BZ - 1506200 - cephmetrics.gauge.ceph in json files
• BZ - 1506239 - [RGW-Multisite] - Metadata sync is not happening from primary to secondary.
• BZ - 1506625 - CVE-2017-12155 allow ceph-ansible to set permissions and then ACL of ceph keyrings
• BZ - 1507128 - rgw_file: objects have wrong accounted_size
• BZ - 1507650 - ceph_test_cls_log failure
• BZ - 1507907 - [ceph-ansible] - osd_auto_discovery set to true is getting ignored
• BZ - 1508615 - [RH Ceph3.0] list bucket which enable versioning get wrong result when user marker
• BZ - 1509035 - On the NFS mount, sub-directories getting created using 'cd' if a similar named child-directory is present.
• BZ - 1510470 - Containerized OSDs don't start - fail to find the Journal device
• BZ - 1510906 - [ceph-ansible] [ceph-container] : conf file generation failing - list object has no attribute u'ipv4'
• BZ - 1511593 - ship python-crypto in RHCEPH 3
• BZ - 1511811 - [Ceph-Ansible]: Upgrading cluster from 2.4 to 3.0 fails for missing mgr keyring.
• BZ - 1511901 - [Ceph-ansible][Ubuntu] Ansible is failing looking for loop devices when osd_auto_discovery is true
• BZ - 1512092 - radosgw segfault after running 'radosgw-admin data sync init'
• BZ - 1512538 - [ceph-ansible] Rhel cluster deployment fails with osd_scenario : lvm
• BZ - 1512594 - [ceph-ansible] [ceph-container] : automatic prepare osd disk failing - non-zero return code
• BZ - 1514460 - ceph-ansible 3.0.{12,13} fails deployment (3.0.11 does not) :: ceph-osd : prepare ceph "filestore" containerized osd disk(s) non-collocated

CVEs

(none)

References

(none)