fix typo in changelog

• Files Changed
• Browse Source

- Version 0.9.40
- Take drivegroup targets into account for rebuild.node (bsc#1198929)
- Allow rebuild.node on empty clusters with no PGs

• Files Changed
• Browse Source

- Version: 0.9.39
- Update ceph.upgrade.ses7.adopt for SES 7.1 (bsc#1197503)

• Files Changed
• Browse Source

- Version: 0.9.38
- Show extra info when PGs take too long to be in active+clean state (bsc#1194807)
- Zap disks when there are no partitions (bsc#1194807)
- Ensure upgrade.status lists nodes with non-ceph roles (bsc#1195366)

• Files Changed
• Browse Source

- Version: 0.9.37
- Bump OSD available timeout to 10mins, make it configurable (bsc#1192520)

• Files Changed
• Browse Source

- Version: 0.9.36
- Allow embedding of Grafana graphs (bsc#1186131)
- Wait for OSDs to be active after restarting (bsc#1185422)
- qa: functests/1node/restart: fix changed.any check
- mds/restart: only check MDS processes (not all processes)
- dg: include unavailable disks (bsc#1181725)
- upgrade: Add ability to specify registry credentials (bsc#1181095)
- Fix no UCST response after 1 second when validating time server
- osd: handle ceph osd ok-to-stop output from ceph v14.2.22
- monitoring: put node_exporter ARGS all on one line (bsc#1188486)

• Files Changed
• Browse Source

- Version: 0.9.36
- Allow embedding of Grafana graphs (bsc#1186131)
- Wait for OSDs to be active after restarting (bsc#1185422)
- qa: functests/1node/restart: fix changed.any check
- mds/restart: only check MDS processes (not all processes)
- dg: include unavailable disks (bsc#1181725)
- upgrade: Add ability to specify registry credentials (bsc#1181095)
- Fix no UCST response after 1 second when validating time server
- osd: handle ceph osd ok-to-stop output from ceph v14.2.22

• Files Changed
• Browse Source

removing 0.9.35 tarball

• Files Changed
• Browse Source

- Version: 0.9.35
- osd: add method to zap simple osds (bsc#1178657, bsc#1178216)
- upgrade to cephadm: fix Drive Group generation (bsc#1181665)
- Rework config change detection to handle global.conf correctly (bsc#1181183)
- Replace yaml.load() with yaml.safe_load()
- Use -i to pass credentials to `ceph dashboard` commands (bsc#1183600)

• Files Changed
• Browse Source

- Version: 0.9.34
- Use correct pool when checking for rados config object
- spec: declare incompatibility with ceph < 14.2.13
- rgw/dashboard: use rgw_frontend_ssl_port in SSL scenario
  (bsc#1178355)
- Fix setting custom Grafana certificates and Prometheus CLI flags
  (bsc#1179581)
- spec: Make deepsea-cli require deepsea
- ganesha: use ganesha_config pool in ganesha.report (bsc#1179621)
- monitoring: fix node_exporter overrides in prometheus.yml
  (bsc#1179029)
- ceph.conf: don't disable msgr2 (bsc#1180594)
- monitoring: enable custom scrape configurations (bsc#1179029)
- runners/upgrade: drop bootstrap_minion and bootstrap_mon_ip
  (bsc#1180960)
- ceph.updates.salt: restart salt-minion.service with bg=True
  (bsc#1181099)

• Files Changed
• Browse Source

Fix changes file entry

• Files Changed
• Browse Source

• Files Changed
• Browse Source

- Version: 0.9.32
- drop workarounds for old ceph-volume lvm batch command

• Files Changed
• Browse Source

- Version: 0.9.32
- Pass --container-init to `cephadm adopt` (bsc#1177588)
- runners/upgrade: Add SES6->7 pre-upgrade checksss --container-init to `cephadm adopt` (bsc#1177588)
- mds: add rescind for multiple mds per host (bsc#1164142)
- mds: add option to manage multiple MDS per host (bsc#1164142)
- Version: 0.9.31
- runners/upgrade: Add SES6->7 pre-upgrade checks
- runners/upgrade: Add generate_service_specs function
- Add ceph.upgrade.ses7.adopt state
- spec: Add /srv/salt/ceph/crash/files directory
- runners/upgrade: Add ceph_salt_config function
- Fix case where remove_subnet_list is None
- validate: check if domain grain is set
- Add dedicated ganesha_config pool for dashboard (bsc#1172689)
- fix many-to-many not allowed in Grafana dashboard (bsc#1175585)
- aa-teardown can fail if apparmor is disabled on the boot command line  
- Create and distribute keyring for ceph-crash (bsc#1174004)
- ceph.conf: remove the mon_health_preluminous_compat settings (bsc#1174545)
- Ignore disks with GPT headers or unavailable 
- Sync salt dynamic modules (bsc#1170752)
- Only remove tuned when a minion has no roles (bsc#1170491)
- Do not remove ceph-grafana-dashboards on mgr nodes (bsc#1168359)
- Version: 0.9.30
- osd: forcefully mark an osd as down after stopping the service (bsc#1171451)
- Version: 0.9.29
- Use salt's stack.py (bsc#1167571)
- Version: 0.9.28
- Correct spelling typo (bsc#1154489)
- Validate iperf results from modules (bsc#1164869)
- Do not rely on isinstance (bsc#1164571)
- Create disabled files to simplify removal instructions (SES-769)
- Fix iSCSI default values
- Grains are not used by ceph-volume (bsc#1160172)
- cephdisks: exclude mdraid devices (bsc#1158196)
- Add upgrade check for `straw` algorithm (SES-356)
- use non-blocking i/o with .wait (bsc#1158184)
- Version: 0.9.27
- cephdisks: add tests for raid config patch 
- cephdisks: exclude disks that are assigned for raid configurations (bsc#1158196)
- osd/replace: only access metadata if present (bsc#1157469)
- Allow for graceful rerunning when removing the cephfs configuration
- Allow custom grafana endpoint (bsc#1154518)
- Version: 0.9.26
- srv/salt/ceph/purge: Changed order of runner calls (bsc#1150168)
- Simplify management of tuned (bsc#1142416)
- Split output to console and Salt log (SES-813)
- Added data tag in cephdisks/dg for filestore OSDs (bsc#1145600)
- Remove recursion, add custom retries (bsc#1154096)
- Evacutate OSDs simultaneously when removing OSDs (bsc#1154096)
- 'make pyc' target should fail if unable to compile
- 'rebuild' functional tests require 3 nodes
- Fix debug line spelling mistake
- Fix error msg handling during salt-run rebuild.node (bsc#1141282)
- Version: 0.9.25
- lrbd to ceph-iscsi upgrade fixes (bsc#1145617, bsc#1145618)
- ganesha: update ceph.conf in stage 3 instead of 4 (bsc#1150643)
- Add status function to advise on upgrade order
- iscsi: SSL enabled deployment by default (bsc#1134536)
- stage 0: don't show failure in CLI when checking for missing packages (bsc#1137973)
- ganesha: don't run the upgrade on fresh nodes (bsc#1148405)
- Add remove option to the ping to remove multiple ip interface
- iscsi: fix iscsi restart before upgrade to ceph-iscsi (bsc#1136066)
- ganesha: report runner to give support info about nfs-ganesha
- Prevent IOLoop error by calling Salt modules correctly (bnc#1123257)
- Handle the apparmor disabled at boot scenario (bnc#1149405)
- Version: 0.9.24
- Added data tag in cephdisks/dg for filestore OSDs (bsc#1145600)
- conf: rgw: drop rgw dns name configurable
- Evacutate OSDs simultaneously when removing OSDs (bsc#1141285)
- flip logic for /available/ disks filtering
- Version: 0.9.23
- fix for .devices access on Devices object
- remove/replace: add path to status report for better identification (bsc#1142663)
- tests: adapt tests for literal wal/db/journal sizes
- dg: allow literal wal/db/journal sizes
- tests: add lower case test params
- dg: allow lower case sizing units
- The disk_led runner crashes (bsc#1142821)
- makefile: update for ubuntu/python 3
- fix(makefile): remove hardcoded user:group
- fix(makefile): update for centos/python3
- dg: tie lvm.osd detection to block type
- monitoring: alertmanager sysconfig needs --cluster.listen-address (bsc#1141956)
- re-work include_unavailable
- add tests for cephdisks.py
- dg: Fix non-lvm osd detection
- cephdisks: ignore sr & rbd devices (bsc#1137587)
- tests: adapt and add tests for dg/osd
- osd: move discovery code to dg.py (bsc#1135179)
- dg: refactor public functions
- cephdisks: restructure public funcs
- disks: adapt to func-rename
- tests: add tests for /dev/mapper and size exclusion
- dg: exclude /dev/mapper/ disks from c-v output (bsc#1141503)
- mds: numerical mds names are forbidden (have been deprecated since 2014) (bsc#1138442)
- populate: don't overwrite cluster config by default (bsc#1140932)
- dg: exclude disks < 5GB (bsc#1141367)
- try to simulate failure for dg bug
- tests: add testcases for osd_per_device
- dg: add osds_per_device to specs (bsc#1141368)
- Fix subvolume mount for generic device or label
- tests: adapt to new pytest syntax (pytest.raises)
- apparmor: ignore aa-disabled return code (bsc#1130930)
- disable the automated upgrade (bsc#1135913)
- Include missing README (bsc#1138325)
- rescind/ntp: remove the correct RPM
- Version: 0.9.22
- cephprocesses: do not exit on missing pillar struct (bsc#1136818)
- apparmor: include apparmor in stage 0, teardown by default (bsc#1137182)
- monitoring: polulate alertmanager peers must run on salt master (bsc#1137400)
- restarts: if no fs is present, restart mds without coordination (bsc#1110528)
- Simplify subvolume creation, add validation (bnc#1135821)
- dg: adapt default drive group to only match 'roles:storage'
- Version: 0.9.21
- doc: generalize info string
- dg: s/wal_devices/db_devices/ to conform guides
- apparmor: fix mgr profile after nautilus changes (bsc#1130630)
- osd: do not assign value to property
- build: remove tuned from the Requires list
- Revert "spec: add explicit prometheus, grafana runtime deps" (bsc#1134058)
- discovery: do not format if salt returned an error
- monitoring: iterate over pillar yaml to generate prometheus yaml
- Version: 0.9.20
- common: install python3-netaddr package
- iscsi: use FQDN in gateways hostnames
- Add parameter to bypass the pillar_check of dg.py(bsc#1135340)
- osd: add independent osd discovery method (bsc#1135340)
- dg: fix filestore deployment (bsc#1135301)
- fix return type for filestore deployments (bsc#1135302)
- dg: make error conditions consistent; don't execute a error message (bsc#1135301)
- iscsi: fix gateway URL when using IPv6 address
- typo: s/to to/to
- Deploy Prometheus Alertmanager SNMP receiver
- srv/pillar/ceph/README should be installed
- Version: 0.9.19
- Add functests for rebuild, fix helptext for rebuild.py
- spec: add explicit prometheus, grafana runtime deps
- select.one_minion: handle case when no minions match criteria
- iscsi: use deepsea wildcard SSL certificate
- grafana: use deepsea wildcard SSL certificate
- dashboard: use deepsea wildcard SSL certificate
- SSL certificate management state
- core: workaround ceph-volume's wal/db VG spanning issue  …
- mds: move reset/shrink-mds-cluster
- Implement rebuild runner
- modules: expr_form has been deprecated over tgt_type
- iscsi: only wait for iscsi gateway if it's ceph-iscsi
- iscsi: set ceph-iscsi logger level to WARNING
- ipv6: fix iscsi configuration in ceph-dashboard
- iSCSI upgrade implementation
- ganesha: use deepsea.state_apply_if to decide ganesha restarts
- ganesha_upgrade: fix call to pkg.info_installed
- select: sort minion list to maintain the order across calls
- deepsea state module - state_apply_if implementation
- Improve RGW configuration in ceph-dasboard
- rgw: allow the configuration of radosgw frontend port using the salt pillar
- Configure grafana in ceph-dashboard
- tests: fix igw test case in cephprocesses module test
- ganesha: upgrade orchestration
- Use correct config name for Dashboard SSL port
- mgr_orch: return ganesha and iscsi endpoint URLs
- core: add wal/db size flag, enable limit: for wal/db devices
- monitoring: fix file_sd_config paths in prometheus config
- monitoring: deploy alertmanager in stage 3
- monitoring: remove old state files
- Use short hostname from salt grains to feed ceph c_v properly (`ceph-volume lvm batch`)
- monitoring: install default Ceph alerts
- core: add include_unavailable flag for disks.py
- tests: add tests for dg.py filestore implementation
- apparmor: add teardown state
- apparmor: teardown existing apparmor by default
- runners/deepsea.py: fix deepsea RPM version detection
- tox: use python3 instead of python3.6
- Remove obsolete includes.
- dg: use re.search instead of re.match
- ganesha: keep generating cephx keys with same name as in ses5
- cephprocesses: don't wait for rbd-target-api because it might not exist
- cli: fix step long descriptions format
- build: order Requires alphabetically
- Update srv/modules/runners/osd.py
- core: add filestore support for dg.py
- build: make drive_group.yml 'noreplace'
- osd.remove to accept compound targets
- chrony: allow NTP access from local network
- Version: 0.9.18
- core: do not skip disks that are not used by ceph
- ceph-volume: use re to extract json string from c-v output
- iscsi: disabled SSL by default
- iscsi: SSL support for ceph-iscsi deployment
- do no parse ceph output when no valid json
- disks runner: guard destroyed() against roles:master being absent
- monitoring: add init -> default pattern to prometheus state
- Avoid running drivegroup based osd deployment post-upgrade
- monitoring/prometheus: add support for user provided rule files
- monitoring: remove old alertmanager state
- monitoring/prometheus: put scrape configs under scrape_configs subdir
- monitoring: prometheus scrapes grafana via https
- monitoring: add alertmanager service discovery to prometheus config
- monitoring: add option to pass flags to prometheus
- monitoring: deploy grafana with ssl by default
- monitoring: add alertmanager rescind
- monitoring: add alertmanager install, service and config
- install ceph in packages.common
- Fix typo: URL reference to Drive-Groups wiki
- rpm: declare runtime dependencies in the spec file
- monitoring: set grafana default theme to light
- Makefile: make "make rpm" less likely to produce a wrong version number
- monitoring: update smartmon.sh, fix label dimension inconsistency
- crc_method: deepsea_minions can be a compound target
- dashboard: support for enabling/disabling SSL and set certificates
- Version: 0.9.17
- rgw: start making beast as a default frontend
- adapt remove and replace logic to ceph-volume + tests
- mgr_orch: return raw ceph_volume output in get_inventory()
- populate: create storage roles for all nodes
- iscsi: use iSCSi gateway public IP address in dashboard configuration
- Remove RGW demo bucket creation
- ganesha: find pool during stage execution (instead of in parsing phase)
- Support IPv6
- Version: 0.9.16
- qa: disable apparmor at end of functest
- ganesha: fix radosgw keyring for use by nfs-ganesha
- ganesha: store ganesha configuration in RADOS pool
- radosgw: automatic configure ceph-dashboard
- orchestrator: disable orchestrator for now
- spec: fix igw file/dir list
- packages: added "curl" to common
- iscsi: wait for ceph-iscsi API to be available
- iscsi: fix service restart
- iscsi: fix gateway restart
- iscsi: fixes several bugs
- iscsi: add igw minions to dashboard
- iscsi: make rbd_pool function more generic
- iscsi: find or create rbd pool to store ceph-iscsi configuration
- igw: iSCSI gateway deployment based on ceph-iscsi
- runner: select: allow specify list of roles in public_addresses
- stage 3: configure deepsea mgr orchestrator module
- cli: fix function args string formating
- Stage 4: add missing failhards to radosgw/core states
- populate runner: fix typo "cepg"
- functests/1node: disable broken symlink test
- cli: automatically wrap of step lines
- Makefile: install symlink tests and functests
- cli: terminal_outputter: fix simple output when printing state exception
- cli: fix stdout/stderr redirection for python 2
- Update CHANGELOG
- cli: fix simple outputter when printing stage exception
- prometheus: fix populate_mgr_scrape_configs states to work with deepsea-cli
- cli: handles render failure when minion does not respond
- ceph.conf: disable msgr2
- Add runner to enable/disable storage enclosure LED.
- Version: 0.9.15
- Add dashboard orchestration
- monitoring/grafana: add config for mgr dashboard
- monitoring: partition node_exporter targets
- stage 3: wait for mgr to be available after it's initially deployed
- Handle broken symlinks
- Install the ceph-mgr-dashboard on mgr nodes
- monitoring: defer deployment to stage 3
- Support ceph-iscsi
- disable restart.mgr tests
- Review functests
- populate should overwrite the stack/default tree at all times
- Add ceph.salt.crc to stage.0
- force 'adler32' as CRC method
- Version: 0.9.14
- Assign drives based on filter rules (SES-150)
- This patch completely changes the way we define osd layouts.
- Please refer to https://github.com/SUSE/DeepSea/wiki/Drive-Groups for documentation.
- monitoring: avoid failed state if no monitoring roles are deployed
- monitoring: simple monitoring restarts
- monitoring: add prometheus and grafana to cephprocesses
- monitoring: add prometheus and grafana rescind
- monitoring: setup and provision grafana with upstream dashboards
- monitoring: deploy to prometheus roles
- monitoring: restarts, avoid "matched no target"
- This patchset add two new roles. "grafana" and "prometheus"
- Version: 0.9.13
- ganesha: drop nfs-ganesha-utils install
- install_ganesha: failhard
- Version: 0.9.12
- Fixes for upgrade procedure
- Check for digit before 'p'
- Makefile: pass --long option to git describe
- packages/common: Fix typo "realease"
- Add retry logic to readlink for seemingly stubborn systems
- Fix lint errors.
- Move pytest to base_python 3.6 and adapt tests accordingly
- Version: 0.9.11
- Save and restore reweight in addition to weight
- look for values of test.ping returns rather only the existence of the key
- Fix auth caps for openstack.integrate
- Fix unicode output error
- Version 0.9.10
- tuned: disable dynamic tuning for all profiles                                                                                                                                                                                                                                   │············································
- ceph-volume dummy: translate by-id device paths to short device paths                                                                                                                                                                                                            │············································
- Change default timeout to 1 hour for redeploying OSDs                                                                                                                                                                                                                            │············································
- Remove populate.engulf_existing_cluster                                                                                                                                                                                                                                          │············································
- Harden split_partition function to do not fail in such case:
- Version 0.9.9
- Makefile: Install python-configobj
- Add mgr_orch runner to support ceph-mgr deepsea orchestrator module
- tests: don't call helper_specs fixture directly
- populate/engulf: don't add cluster-unassigned line
- Return version in cephdisk.version() instead of printing it. So ``salt '*' cephdisks.version`` will output correct values instead of None.
- Consider iptables not installed as passing
- Check configs for deprecated k:v pairs and report
- build/ops: install python3-prometheus-client
- osd.py: adjust pg_states() method to Nautilus
- remove traces of oA
- qa: disable engulf functional test (not working with LVM)
- Makefile: Fix two missing $DESTDIRs
- Makefile: clean up suse conditional blocks
- Prevent rgw-ssl name collision when pushing certificates and using custom rgw roles
- Version 0.9.8
- dummy lvm: add basic ceph-volume implementation and use it
- Pin the linter to a pre-change version
- lrbd service restart instead of reload on `ui_iscsi.deploy`
- Correct conditional check for custom configurations
- Makefile: Check for correct ID in /etc/os-release
- Add more info in the warning/error validators
- Add repo refresh check
- Add validation for pending updates
- cli: prettyprint: force stdout flush in all stdout writes
- Makefile: drop redundant PIPCMD
- Add documentation to manpage and trim whitespaces in existing pages
- Refactor cephprocesses module and add tests for it
- fix: only check OSDs after restarting an OSD
- Add cluster shutdown, startup functionality
- Remove hardcoded nvme devices, support generic convention
- Support forced restart
- mds/restart: fix restarts when minion_id is fqdn
- cli: terminal_outputter: don't dereference a runner step that already finished
- cli: stage_parser: fix state rendering when no minions match the target
- qa: add baseline engulf functional test
- remove: wolffish from master
- populate/engulf: support optional master_minion
- cephinspector: Fix typo preventing OSD detection
- populate/engulf: fix local variable in _get_existing_cluster_networks()
- populate/engulf: use ConfigParser.read_string() instead of StringIO
- populate/engulf: overwrite profile-import during engulf
- populate/engulf: use .format for string manipulation
- populate/engulf: unify returns and error messages
- populate/engulf: print info about engulfed config
- Remove journal_size if present
- Version 0.9.7
- build: Add %pretrans script to deal with dir->symlink change
- qa: drop health-ok.sh and associated scripting
- Specify only support image feature: layering
- Install salt-api requirement prior to building the package. Trying to fix #1285
- Remove error condition, empty cluster network is valid
- man: missing closing quotes
- build: Remove duplicate functests `install` lines
- qa/deploy: fix error-reporting issues
- functests: fix pid comparisons when process not running
- functests: stop restart test if no pid
- cli: refer to orchestration instead of stage
- qa/common: properly report stage timeout when running via cli
- qa/health-ok: repeat Stage 0 only once
- qa/deploy: use jq instead of python in salt_api_test
- qa/common: move salt_api_test to deploy.sh where it is used
- cli: fix parsing of salt.runner when used inside state files
- remove/replace interference fixes
- man: add line break between separate commands.
- fixed problem in stage.0 with error message no  minion found in metapackage
- cli: support list of sls ids in salt.state orch function
- qa/health-ok: add missing options to usage message
- qa/deploy: create rbd pool if --rbd option passed
- qa: move pgs_per_pool function to common/pool.sh
- qa/health-ok: implement --rbd option
- qa/pool: create non-prefixed openstack pools
- build/ops: make "make rpm" work in a py3-only environment
- qa: increase stage timeout duration from 30m to 60m
- qa/health-ok: add --tuned=on/off option (default: off)
- qa: add openstack to ceph.functests.1node orchestration
- qa/health-ok.sh: add --openstack option
- s/unsused/unused/
- qa: don't initially delete openstack pools
- monitoring: fix fs fillup time query
- tuned: make stanza names unique
- monitoring: fix rbd_exporter pool parsing for ids > 9
- Set theme jekyll-theme-cayman
- Use disks paths closer to real hardware for tests
- Refactor tests for test_proposal.py
- Remove codecov from unit tests
- NFS-Ganesha + Cephfs: Add cache config options
- Add test for OSD migration (different slots)
- Add OSD migrations that include disk slot changes
- Restart is separate state now
- Add wiki to PULLREQUEST_TEMPLATE
- qa/policy: MDS on up to 3 storage nodes only
- qa/health-ok: drop --fsal option (no longer needed)
- Return devicename if no /dev/disk/by-id path
- qa/health-ok: implement --no-update option
- qa: more flexible role assignment in policy.cfg
- stage/prep: reinstate single quotes around deepsea_minions tgt
- qa/health-ok: use "echo" instead of "echo -n"
- qa/health-ok: add NFS-Ganesha deployment/test functionality
- qa: health-nfs-ganesha.sh fix case statement
- qa/rgw: restart rsyslog service after installing curl RPM
- qa: use "systemctl --no-pager --full status"
- cephinspector: use cephdisks.device in engulf (bsc#1082994)
- cephinspector: assume filestore OSD if type file not present (bsc#1104773)
- Remove deprecated test function
- Moved stage 0 specific man sections to the correct file, as per review comments
- Make jinja variables quoting consistent
- Change minion to default update-no-reboot
- Update man pages of stage 0 and customization to update-no-reboot
- Change stage 0 default toupdate-no-reboot
- rescinded.py: tweak runner comments
- Add unittests for wait.just
- doc: drop osd.tree command
- qa: restart ceph-radosgw service after installing curl RPM
- Stage 5: delete CRUSH host buckets orphaned when storage node is deleted
- rescinded.py: implement rescinded.delete_orphaned_host_buckets
- osd.py: refactor osd.tree
- qa: common/deploy: run "zypper ps -s" on all nodes
- populate: don't create role-mon yml line during engulf (bsc#1104781)
- Use host rather than nodename consistently
- Reuse OSD ID in migrations, support replace.osds
- common/helper: show output of test.ping in _first_storage_only_node
- common/helper: storage-only node might be master node
- Refactor device detection, allow override
- Use fqdn for default minion name in template
- qa: run lsblk and ceph-disk on storage node only
- Explicitly add hwinfo and lshw
- qa/deploy: handle possibility that deepsea-cli is not installed
- Add checklist to Pullrequest template
- qa: add script to test Stage 5
- Fix indentation for tests/
- fix readme matrix rendering
- qa: health-ok.sh: enable testing of OSD configurations
- rescind: prevent empty rendering of ceph.rescind.tuned state file
- build/ops: streamline qa %files section
- cli: filter out test.nop steps and don't report [init] step errors
- qa/common: create btrfs subvolumes after Stage 0
- qa/common/helper: unify whitespace
- qa/health-ok: fix usage message (--encryption was missing)
- qa/common: unify whitespace
- qa: Add testing osd-configrations qa: add README to clarify
- qa/common/deploy.sh: fix redirect of stderr to /dev/null
- qa: rename ceph.smoketests to ceph.functests.1node
- qa/health-ok: add --min-nodes and --client-nodes options
- qa/common/helper.sh: refactor _run_stage
- qa: health-ok.sh: add a top-secret undocumented navigation code
- qa: fix ceph_version_test
- qa: make MDS deploy even without client nodes
- qa/common/helper.sh: drop workaround for bsc#1087232
- qa: refactor health-ok.sh, drop health-{mds,rgw}.sh
- apparmor: install apparmor-utils package on SUSE
- openstack: add smoketests for ceph.openstack
- openstack: add openstack.integrate runner (fate#324708)
- openstack: allow specifying prefix, use generic pool names
- openstack: enable rbd application on openstack pools
- openstack: add cinder-backup user and pool
- openstack: collapse keyring cache directories
- openstack: don't create separate key for nova service
- add test for kernel validation
- add rbd kernel module validation
- build/ops: add /srv/salt/ceph/smoketests/init.sls to spec file
- build/ops: add /srv/salt/ceph/smoketests/init.sls to Makefile
- Rename tuned profiles from ses-* to ceph-*
- Only remove tuned if a node has no osd, mon or mgr
- Add tuned smoke tests
- Add predefined tuned profiles
- enforce pylint version 1.9.x
- build/ops: add jq runtime dependency to spec file
- fix issue that would make benchmark.baseline fail if an osd is removed. make benchmark.baseline use 'ceph osd ls' instead of 'osd.list' to get the list of the active osd ids.
- address suggestions & fix typos
- Adding more information to the man pages
- adaptions to deepsea-qa package
- install smartmontools in stage.0 (bsc#1104779)
- Do not create untracked partitions
- rescind: prevent empty rendering
- fix function header after adaption
- use safe_dump rather than dump (bsc#1100136)
- Version 0.9.6
- Fix uuid, restart, lint issues, Makefile
- Fix migration of empty cluster, salt api validation
- Cleanup, remove uncalled code
- Improve ceph version check
- Add mon/mgr check after start
- Support apparmor
- Lock tox to Salt 2016
- Address FIXMEs, use jq, bare except
- Fix cli for absent decoding
- Address Salt cache permissions
- Support metapackage
- Experiment with py-coverage
- Add iputils dependency
- Version 0.9.5
- Master minion setting is optional
- Add restart smoketests
- Add Mimic compatible OSD bootstrap
- Add os_switch macro
- Fix mds restart for multiple active mds
- Add validation for Salt version
- QA refactor, check ceph-test is installed
- Fix cached runners
- Fix makefile for Fedora
- Contributing doc update
- Version 0.9.4
- Require Ceph Luminous or newer
- Allow custom format for select.minions
- Remove filestore_xattr_use_omap
- Support kwargs for proposal.populate
- Fix validate conditionals, redeploy nodes init.sls, apparmor command
- Cleanup remove unused OSDState, remove obsolete public address method
- Correct Jinja conditional for lrbd.conf.j2
- Add active+clean, safe-to-destroy checks
- Add keyring smoketests
- Add rescind ceph configuration
- Prefer detailed device names for OSDs
- Refactor changed.py, fix arg name, support igw restart
- Fix engulf to return single device
- Support python3, rework imports
- Support Salt 2018
- Change default PG  ratio for Cephfs pools to 4:1
- Workaround for job cache permissions
- QA: Add --mini for health-ok
- Version 0.9.3
- PR template
- Fix node_exporter
- Create cli subpackage
- Fix another wal/db check
- Support multiple networks in engulf
- Tune chrony
- Additional qa functions, migration testcase
- Version 0.9.2
- Stage 4 restart complete
- Support prometheus 0.15 node-exporter
- Fixes for device discovery, wal/db checks
- Switch to chrony
- Add s3 checks
- Many QA additions
- Add blockdev benchmark
- cleanup, remove diagnose files
- Add editorconfig
- Swap salt-api and sync order
- Improve policy.cfg error handling
- Version 0.9.1
- Fix ceph package install instead of latest
- Change spec file version management
- Initial Openstack support
- Initial CentOS support
- Version 0.9
- Refresh package db
- Reimplement restart orchestrations for config and binary changes
- Support restart differences between upgrade, Stage 0 and Stage 3
- Update man pages
- Improve rgw.add_users
- Begin benchmark-rbd role
- Optimize bluestore argument handling
- Support multiple ntp servers
- Grafana dashboard touch ups, include /var
- Lint fixes, enforce via Travis
- Correct behavior for failed engulf
- Import Validate fix
- Add sortbitwise flag
- Use host grain for shortname, Crowbar compatibility
- Use pyfakefs 3.2 for unit tests
- Fix is_prepared for NVMe OSD
- Include openATTIC, client processes in cephprocesses
- Use mine.update instead of mine.restart
- ui_iscsi fixes
- Fix radosgw system user creation
- Add selectable cron for radosgw-admin usage
- Suppress ceph-disk deprecation warning
- Support individual iSCSI gateways
- Add version to Salt api
- Fix OSD drain
- QA restart tests, display rpm version
- QA consolidate encryption, validate rgw system user
- Version 0.7.35
- Relocate openATTIC config file
- Fix cli
- Version 0.7.34
- Fix Makefile, spec file
- Version 0.7.33
- Update grains unconditionally
- Version 0.7.32
- Cleanup: package removal
- Add argument to benchmark runner
- Additional qa tests
- Version 0.7.31
- Moved salt-api validation
- fix NVMe partition removal
- Version 0.7.30
- Swap salt-api and validate.setup
- Fix syntax indentation
- Version 0.7.29
- Minor dashboard changes
- Subvolume support
- New man pages, help functions
- Fix cephprocesses, cephinspector for psutil
- Cleanup error messages, move pillar examples to doc
- Fix cephdisks for smartctl detection
- Version 0.7.28
- Fix ganesha config parsing
- cli: use ascii on unsupported UTF-8 terminals
- cli: render/substate/privilege fixes, run as salt user
- grafana dashboard corrections, top 10 stats
- remove unused iperf directory
- select.minions support multiple values
- restart RGW gateways only when config changes
- rgw: bucket query, demo bucket restricted to demo users
- rgw: ssl cert permissions
- fix validate for DEV_ENV in pillar for single node
- monitoring: cron job lock, restrict to fast-diff images
- improve validate openATTIC/RGW port conflict check
- support rgw_configurations in cephprocesses module
- enable rgw usage log
- include cron package for Prometheus
- set aio-max-nr, fixes OSD creation on larger nodes
- another deepsea_minions example
- Version 0.7.27
- Improved RGW dashboards
- deepsea command for viewing orchestrations
- Minor fixes in populate runner
- Separate RGW users, add system user
- benchmark improvements
- fix oA, RGW port validation
- fix sort order for net.iperf
- qa tests for encrypted disks, oA/RGW deployment
- Version 0.7.26
- Handle missing instance of ceph.conf for rgw
- Version 0.7.25
- Fix net.iperf
- Fix ceph.conf for rgw
- Version 0.7.24
- Fix ceph.conf generation
- Fix proposal runner
- Version 0.7.23
- Support Bluestore on importing non-DeepSea clusters
- Add deepsea_minions validation to engulf
- Improve cluster network detection in engulf
- Add pyc files to rpm
- Create qa subpackage
- Fix Stage 5 storage node removal
- Reimplement mon_host and mon_initial_members
- Fixes for test environment
- Notify admin about public, cluster setting
- Minor corrections, updates to monitoring, runners
- Version 0.7.22
- Rename ceph_tgt to deepsea_minions
- Minor promethus fix, add ganesha role to policy example
- Version 0.7.21
- Monitoring updates
- Revert rgw.sls removal from /srv/pillar/ceph/init.sls
- Add python-rados dependency to spec
- Version 0.7.20
- Update README for DEV_ENV
- Various Grafana dashboard fixes/improvements
- Add qa test cases for ganesha, sanity checks
- Support ssl for rgw configurations
- Add mandatory ceph_tgt constraint
- Enable salt-api, include validation
- Remove unused client roles
- Version 0.7.19
- Revert salt-api
- Update grafana dashboards
- Fix proposal runner
- Version 0.7.18
- Enable salt-api
- Add rgw-ssl state, examples
- Fix upgrade check on minons.ready, remove icinga references
- Add deepsea.commands man page
- Move openattic example to rescind/openattic
- Enable iscsi-images pool
- Add module for configuring openATTIC, mds caps to keyring
- Check for openATTIC and rgw role conflicts
- Relax firewall check, allow custom chains
- Fix apparmor check for importing clusters
- Version 0.7.17
- Change default proposal to new profiles
- Propose bluestore OSDs by default
- Fixes for upgrade orchestration (bsc#1043863, bsc#1048186)
- Fixes for import of ceph-deploy cluster
- Add support for new 'ceph health' format
- Add smartmon and rbd monitoring to grafana
- Add mgr removal (bsc#1047892)
- Fix rbd demo image creation for iscsi
- Deploy salt-api and add eauth sharedsecret (bsc#1049215)
- Fix mgr caps in various keyrings (bsc#1049151)
- Improve integrtation tests
- Update benchmark capabilities
- Version 0.7.16
- Upgrade fixes
- Support importing ceph-deploy clusters
- Bluestore no longer experimental
- Add mgr role, keyring fixes
- Add ceph.conf.rbd
- Add librbd testing
- Move demo image to iSCSI stage
- Change Stage 5 default to not remove pools
- Add alternate Stage 0 states for no reboot/update
- Improve filestore to bluestore migration, multiple wait states
- Migrate policy.cfg and hardware profiles
- Cleanup, remove unnecessary sls files
- Version 0.7.15 8172280bf2ad29584577d0d1ce256a82ff3009cf
- Fix reweight, remove strays
- proposal runner fixes
- Version 0.7.14 fc9b64b6778db158bca08072dde3b8c1df8b01fe
- Fixed bug in RGW deployment
- Version 0.7.13 0e17664447ed75cecbfbc7bf5e7051a957c56fc4
- Progressing with integration test suite
- Initial support for importing ceph-deploy clusters
- Correct rgw endpoints, several ganesha configuration changes
- Default Grafana desktop fixed
- Migration by OSD working
- Many minor fixes
- fix potential race between partprobe and dd
- Version 0.7.12
- re-enable monitoring deployment
- add initial functional testing suite
- Version 0.7.11.1
- do not deploy any monitoring be default, avoid MON regression in 12.0.2
- Version 0.7.11 2ffc279ae43724e1b0e349d0a967bf7ee83c39e9
- Add prometheus, grafana and basic configuration
- Improve OSD deployment
- Add remove functionality and single OSD removal command
- Ganesha tuning, add ganesha runner for salt-api
- Various spec file, build environment changes
- Version 0.7.10 8173ccabc62729981ccc876966231aee0ba4b9d4-
- Fix is_activate and is_prepared with by-id devices
- Wal and db partuuid awareness
- Fix osds+journal deployments
- Version 0.7.9.1 48c933d0e9405198fe8873e7fe3a61d8cd54fa73-
- Fixed bug in osd module internal targeting
- Fixed conversion of bytes to gigabytes for disk size detection
- Version 0.7.9 0fbdff9abab508277940d50d775fb01bb515860c
- Automatically include custom files from ceph.conf.d
- Enable client ntp configuration
- Add ui_rgw runner for RadosGW credentials
- Add prometheus exporters (prometheus, itself is in progress)
- Add validation to discovery stage, linter for yaml, check profiles
- Ensure at least one storage in DEV_ENV mode
- Fix storage removal order
- Warn if apparmor is enabled
- Enable bluestore, sizes for journal, wal and db
- Remove ganesha /etc/sysconfig/ganesha (moved to rpm)
- Initial dmcrypt support (removal in incomplete)
- Add status.report runner, display after upgrade
- Misc. fixes to READMEs, radosgw rescind
- Version 0.7.8 93ec5f36f4a8c27cc1fac0c841ccf45bab6841bb
- salt-api fix
- Updated README
- Version 0.7.7 8f803eabe3294b205af17786417e26b2143e61f1
- Support restarting single services
- Added openATTIC role
- Rename cherrypy to salt-api
- Add Wolffish to contrib
- Additions to ui_iscsi api
- Fix ganesha service file
- Version 0.7.6 18a3f872ad438332e2aebb130fa5e4a79bcb0bd6
- Rolling upgrade
- Remove unused directories, modules
- Include python bytecode in rpms
- Migrate ganesha to salt modules for restarts
- Correct package dependencies for Ubuntu
- Version 0.7.5 99f5be86c4d5c49c30d29cf68f228beb0adcf5a5
- Fix bugs for ceph.purge, disengage.safety
- Skip unassigned service orchestrations
- Add pylintrc and associated bootstrap script
- Fix ganesha ordering, restart, validation
- Fix permissions, encoding of runners, modules
- Add various unit tests - filequeue, push
- Improve comment handling in policy.cfg
- Add shared keys for mds, rgw
- Correct building, dependencies on OpenSUSE
- Fix certificate of origin, url in contributing.md
- Change Stage 0 ordering
- Support DEV_ENV flag
- Enable openATTIC rpcd, systemd services
- Add cephservices runner, module - renamed cephprocesses
- Fix eauth for cherrypy configuration
- Change cephfs pools initial pg from 256 to 128
- Rewrite cephdisks to handle raid controllers, support lspci
- Support multiple public, cluster networks
- Various python improvements, remove unnecessary methods
- Version 0.6.13 git 709867f2ac1c55931504fb2eea0fb0a6fcb369fb
- Enable openattic services (bnc#1027261)
- Version 0.7.4 e8559976b5e70dd36e9a98250f795de9f1cc48d4
- Add ganesha for rgw/cephfs
- Add kernel.replace
- Add missing import
- Version 0.7.3 316c3b9071807a22baa87df5dd1823560c9ef79e
- Refactoring, code improvements
- Disable unsupported kernel features for rbd images
- Correct time sync bug in Stage 3 when time_service is disabled
- Add cherrypy, salt-api support
- Add net.ping runner, multi.ping module
- Version 0.7.2 0aeeceda3497cd3dad2439954203829ab428f43a
- Add python-ipaddress dependency
- Add unit tests, refactor
- Add warnings to validate
- Version 0.7.1 git 2eaa452cbe6f24538a91433575010e7ef9bf1f2f
- Fix various README typos, add links to documentation
- Add unofficial mascot
- Rename 'name' parameter
- Add alternate default for kernel-default
- Add tox.ini, test structure
- Add disengage.safety, ceph.purge
- Version 0.6.12 git a2ac3b3da52068c085159928716d3bbc4e84eb97
- Do not replace existing /srv/pillar/top.sls (bnc#1020840)
- Version 0.6.11 git 3b8e493eac552ee6b3051fb513704aff2fcbd5e8
- Renamed "name" parameter
- Version 0.6.10 git 73f55a985b62c171ef4d6af29d94dc458cd00733
- Fix ordering of Storage node removal
- Version 0.6.8 git 56a0dbea92640a0212104099cc3fe09c2e6b64d9
- Remove extraneous spaces/comments from runners/modules
- Fix policy.cfg examples, remove old README
- Remove unnecessary defaults from pillar
- Remove stack/default before writing
- Use only first device from hwinfo
- Create pools from master
- Change Salt state_output to mixed
- Version 0.6.7 git d4ea8857fffa1a9da4c19285506266d5ea30f3c9
- Rename custom.sls, add comments and general cleanup
- Version 0.6.6 git f20b1bd040e2f3674c1c255340fa26716cafe972
- Added baseline benchmarks
- Fixed proposals for nice ratio and unused OSDs
- Version 0.6.5 git ad8b6991805a6c2037ad90205ae504330cb29b68
- Add graceful restart functionality
- Version 0.6.4 git f22ee4176ce6b4cb25f33b55fd02df6299f2bddf
- Renamed rescind directories to match client roles
- Version 0.6.3 git 6e416b6a2c68b40bdbb90482a1ac8295670a97fb
- Migrated install from spec to Makefile
- Version 0.6.2 git da5939238bdd5a807f8400f11152601408ef6d44
- Proposal corrections for OSDs with separate journals
- Version 0.6.1 git 0dcb54000c874299d0dfbeae0e286595de34c779
- Rename profile, clients and zypper commands
- Version 0.6 git b5a23522ecf114e90d9fea578e6648b3a04a41f6
- Rename variables from _method to _init
- Rework stage directory to allow overriding
- Add valdiation checks
- Version 0.5.8 git 538dbbdee4b0f16f05ab1ca05087480162595f79
- Create default iSCSI setup
- Restart apache2 for OpenATTIC
- Add iperf/ping diagnostics
- Version 0.5.7 git dd22795643e4785de7f7e6bfc5a76a8bde39b212
- Add validate check for no responding minions
- Finish removal of monitors, iSCSI, CephFS and RGW
- Version 0.5.6 git 7da15965480dcfe901b1814cd077a6e523f8c5ec
- Automatically install openattic
- Add missing directories to spec file
- Version 0.5.5 git 45e2c13ea7159678f1bbc63efdf6d694824776ea
- Add rescind functionality
- Version 0.5.4 git 92b8e8b9dda29aa0a0125ae41b88387055f88a9a
- Correct generation of hardware profiles
- Version 0.5.3 git 3fc096d39d187efd87945cfddd937d561379cdde
- Generate ceph.conf correctly for custom rgw configurations
- Auto accept licenses for updates
- Version 0.5.2 git fa140169e6210ce27cb1164386b3e2fa340b4fa0
- Messages at startup and reboot
- skip CephFS pool creation if no minion has an mds role
- Validate has JSON output option
- Install sgdisk during Stage 0
- Version 0.5.1 git b7465074cb36d304063589db3c9e2b85ebbc0c76
- Handle unassigned roles
- Version 0.5 git 7d541d14d7c8de5b095977d398ffd1d6ea3284f9
- Reimplementation of keyrings, secrets are no longer in pillar
- Version 0.4.5 git 92185470ef471eb07b6afc655b43c5928337b0ba
- Add BuildRequires, remove conflicts
- Version 0.4.4 git 9fffc7e2aa6776c9a7d1bf585fe02f519f1d49a8
- Restart salt-master during installation
- Version 0.4.3 git 1fc618a8c4af50745f2e52d464b32c3a5cd329d3
- Fix spec file permissions and globbing
- Version 0.4.2 git 2b0503d67322788e4490f28268025757a6e797be
- handle missing keys gracefully
- Version 0.4.1 git cfb472439a09ee470b3a95d73d98b952d95e7b3a
- Validate ceph version in every minion repo
- extend README, detail policy.cfg
- Version 0.4 git ca66ab6df47e3ec2d5a4715e1517c4eb568ddce9
- Add validation of master_minion setting in Stage 0
- Version 0.3.1 git 48608e8b6c0fece0c3d1367138c909e1f001ad2d
- Fixed additional misplaced files
- Version 0.3 git 2dd422ae5ff3fbf289f52c390faaa1ddd5712047
- Fixed misplaced files
- Correct LICENSE file
- Version 0.2 git 9eaa8484b53ba1cb2514c8edbb25addfd3451891
- first package

• Files Changed
• Browse Source

- runners/upgrade: Add SES6->7 pre-upgrade checks
- runners/upgrade: Add generate_service_specs function
- Add ceph.upgrade.ses7.adopt state
- spec: Add /srv/salt/ceph/crash/files directory
- runners/upgrade: Add ceph_salt_config function
- Fix case where remove_subnet_list is None
- validate: check if domain grain is set
- Add dedicated ganesha_config pool for dashboard (bsc#1172689)
- fix many-to-many not allowed in Grafana dashboard (bsc#1175585)
- aa-teardown can fail if apparmor is disabled on the boot command line  
- Create and distribute keyring for ceph-crash (bsc#1174004)
- ceph.conf: remove the mon_health_preluminous_compat settings (bsc#1174545)
- Ignore disks with GPT headers or unavailable 
- Sync salt dynamic modules (bsc#1170752)
- Only remove tuned when a minion has no roles (bsc#1170491)
- Do not remove ceph-grafana-dashboards on mgr nodes (bsc#1168359)

• Files Changed
• Browse Source

- Version: 0.9.30
- osd: forcefully mark an osd as down after stopping the service (bsc#1171451)

• Files Changed
• Browse Source

- Version: 0.9.29
- Use salt's stack.py (bsc#1167571)

• Files Changed
• Browse Source

- Version: 0.9.28
- Correct spelling typo (bsc#1154489)
- Validate iperf results from modules (bsc#1164869)
- Do not rely on isinstance (bsc#1164571)
- Create disabled files to simplify removal instructions (SES-769)
- Fix iSCSI default values
- Grains are not used by ceph-volume (bsc#1160172)
- cephdisks: exclude mdraid devices (bsc#1158196)
- Add upgrade check for `straw` algorithm (SES-356)
- use non-blocking i/o with .wait (bsc#1158184)

• Files Changed
• Browse Source

- Version: 0.9.27
- cephdisks: add tests for raid config patch 
- cephdisks: exclude disks that are assigned for raid configurations (bsc#1158196)
- osd/replace: only access metadata if present (bsc#1157469)
- Allow for graceful rerunning when removing the cephfs configuration
- Allow custom grafana endpoint (bsc#1154518)

• Files Changed
• Browse Source

- Version: 0.9.26
- srv/salt/ceph/purge: Changed order of runner calls (bsc#1150168)
- Simplify management of tuned (bsc#1142416)
- Split output to console and Salt log (SES-813)
- Added data tag in cephdisks/dg for filestore OSDs (bsc#1145600)
- Remove recursion, add custom retries (bsc#1154096)
- Evacutate OSDs simultaneously when removing OSDs (bsc#1154096)
- 'make pyc' target should fail if unable to compile
- 'rebuild' functional tests require 3 nodes
- Fix debug line spelling mistake
- Fix error msg handling during salt-run rebuild.node (bsc#1141282)

• Files Changed
• Browse Source