Update skyhook to ceph-nautilus #102
Open
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
After 5ace82e65c72847fb875fc01c419937a26a59d70 was merged, I found three more instances of the code being patched. The commit message of 5ace82e65c72847fb875fc01c419937a26a59d70 was/is: "The default values are handled by mgr_module.py's _get_module_option(); the or here means that we break any non-true (0, false, none) value and override it with the default." Fixes: 5ace82e65c72847fb875fc01c419937a26a59d70 Fixes: https://tracker.ceph.com/issues/43746 Signed-off-by: Nathan Cutler <[email protected]> (cherry picked from commit 30585f44de6147c824ed2df1a477ad17f7b51fff) Conflicts: src/pybind/mgr/osd_support/module.py - file does not exist in nautilus
Additionally, introduce `task status` field in manager report messages to forward status of executing tasks in daemons (e.g., status of executing scrubs in ceph metadata servers). `task status` makes its way upto service map which is then used to display the relevant information in ceph status. Signed-off-by: Venky Shankar <[email protected]> (cherry picked from commit 5c25a018643b10aa78db8270cae1476f71d8f4f4) Conflicts: src/messages/MMgrReport.h src/mgr/DaemonServer.cc src/mgr/ServiceMap.h
Signed-off-by: Venky Shankar <[email protected]> (cherry picked from commit 625dffe65c0f8001b3b6ca6d0b12732a1a103849) Conflicts: src/mds/MDSRank.cc
... also log new and completed scrubs. Signed-off-by: Venky Shankar <[email protected]> (cherry picked from commit 49520156414cb0667e8edd6ea4b31462e7cb7752)
Signed-off-by: Venky Shankar <[email protected]> (cherry picked from commit 465a3adc6c31fd9b8359920ab47adf8e1f45d5f1)
Fixes: http://tracker.ceph.com/issues/36370 Signed-off-by: Venky Shankar <[email protected]> (cherry picked from commit 05d17994a879995b56bda5d770a938d0aabaaed9)
* partially revert 5c25a018. which is not backward compatible. * change `ServiceMap::get_daemon()` so it returns a `pair<Daemon*,bool>`. git is an `optional<uint64_t>`, so we cannot dump it without checking. Fixes: http://tracker.ceph.com/issues/41424 Signed-off-by: Kefu Chai <[email protected]> (cherry picked from commit 3e65551d0ab4acafaf4012601987b3391b2411ed)
Fixes: http://tracker.ceph.com/issues/42169 Introduced-by: 5c25a01864 Signed-off-by: Venky Shankar <[email protected]> (cherry picked from commit cda3eadbfbd3e3a8a057cc7ed042ba0a6d7fef11)
Introduced by 625dffe65c0f8001b3b6ca6d0b12732a1a103849, which added periodic scrub stats reporting. Fixes: https://tracker.ceph.com/issues/42494 Fixes: https://tracker.ceph.com/issues/41525 Signed-off-by: Sage Weil <[email protected]> (cherry picked from commit a1a220d137f7fa128d08f96d1eff83865388e5fc)
This also cleans up the output to be more readable/useful in debug output. Signed-off-by: Patrick Donnelly <[email protected]> (cherry picked from commit 1c56632e88a126d0eac75235d9a9716833cba6b7) Conflicts: PendingReleaseNotes Fix minor conflict in PendingReleaseNotes.
This is a trivial refactor. Signed-off-by: Patrick Donnelly <[email protected]> (cherry picked from commit bafa3b731ff93a356fcf60ebbaf0e5f299182992)
Signed-off-by: Patrick Donnelly <[email protected]> (cherry picked from commit 3355581a298963711ad058c6e47d21b7eab887a4)
Few things here: - Make explicit the check for getting removed from the MDSMap. This was only done before by checking if MDS held a rank which does not check the case where a standby is removed from the FSMap. - Use mds_info_t::dump to simplify various debug output. - Add a few sanity asserts for invalid state transitions. Signed-off-by: Patrick Donnelly <[email protected]> (cherry picked from commit 26a08df2adc7495fc660113cd26c33d5debd3ee6) Conflicts: src/mds/MDSDaemon.h Nautilus still uses "const_ref" rather than "cref_t<>" in master.
The Monitors send an empty MDSMap to an MDS it is removing. The MDS can't diagnose the cause. Instead suggest looking at the cluster/monitor logs. Signed-off-by: Patrick Donnelly <[email protected]> (cherry picked from commit c385178b0c04d09bc6774acf5bdb4537fa6803ad)
This commit undoes the service daemon registration for the MDS. It doesn't look
absolutely necessary and it causes the MDS to be listed twice in the `ceph
versions` output:
$ ceph versions
...
"mds": {
"ceph version v15.0.0-6915-g0143b904676 (0143b9046763ea1801efa8358a0c033ec862cea9) octopus (dev)": 3
},
"mds": {
"unknown": 3
},
"overall": {
"ceph version v15.0.0-6915-g0143b904676 (0143b9046763ea1801efa8358a0c033ec862cea9) octopus (dev)": 10,
"unknown": 3
}
Fixing that requires looking for duplicates or ignoring MDSs in the
service daemons when the mon processes `ceph versions`. I have a feeling
that it wasn't actually designed to be used by the MDS this way however.
Additionally, the reason for "unknown" version is because the metadata
sent to the mgr does not include "ceph_version".
Signed-off-by: Patrick Donnelly <[email protected]>
(cherry picked from commit 046137f819aae72f1423e3feb213f0e46c97c9ce)
Note that we now sub to the mgrmap after init because the MgrClient connection to the mgr is driven by receipt of the MgrMap. This is important so that the MDS does not have metadata with the mgr when the mons are ignoring the MDS otherwise due to CompatSet incompatibilities. Fixes: https://tracker.ceph.com/issues/41538 Fixes: https://tracker.ceph.com/issues/42635 Signed-off-by: Patrick Donnelly <[email protected]> (cherry picked from commit e765f2d533440cfc4189f36fcaba24617a302e84) Conflicts: src/mds/MDSDaemon.cc Nautilus uses Lock/Unlock rather than lock/unlock in master.
This would be widely required since ceph metadata server entries are maintained in service map (DaemonServer::pending_service_map). Such normal ceph services would need to filtered when processing the service map to avoid extraneous entries getting processed. Signed-off-by: Venky Shankar <[email protected]> (cherry picked from commit 79503fc16749ed0cfe8a89ea3b3c8c792d6b8809)
This is done is couple of places in ceph manager -- when culling entries from service map and the other when dumping serice status. Signed-off-by: Venky Shankar <[email protected]> (cherry picked from commit 20139d12423fc6154b57499c576d1e4bb3f1eade) Conflicts: src/mgr/DaemonServer.cc Switch to non-initializer list in range-based for loop.
…tadata Signed-off-by: Venky Shankar <[email protected]> (cherry picked from commit dfc056028e44cdbda84d7118e29120ae70edb35a)
MDS does not register with manager, therefore `task status` (scrub status updates) were not getting displayed in ceph status output. Signed-off-by: Venky Shankar <[email protected]> (cherry picked from commit d6ad0a255a27c5e3da20c7ae03066a692a6d49bb) Conflicts: src/mgr/DaemonServer.cc src/mgr/DaemonServer.h mgr/DaemonServer.{h,cc} deals with raw pointers while master uses ref_t<> cast -- adjust to that. a minor conflict in the header and the metrics templatization is not backported to nautilus. also, DaemonKey is a std::pair in nautilus but a struct in master -- that requires a change in referencing daemon type and name.
Fixes: http://tracker.ceph.com/issues/42835 Signed-off-by: Venky Shankar <[email protected]> (cherry picked from commit fb4cd42cbcdaa4812838b846ae33ee26356ce733)
nautilus: mds: display scrub status in ceph status Reviewed-by: Patrick Donnelly <[email protected]> Reviewed-by: Greg Farnum <[email protected]>
Introduce a config option called 'mon_warn_on_pool_no_redundancy' that is used to show a health warning if any pool in the ceph cluster is configured with a size of 1. The user can mute/unmute the warning using 'ceph health mute/unmute POOL_NO_REDUNDANCY'. Add standalone test to verify warning on setting pool size=1. Set the associated warning to 'false' in ceph.conf.template under qa/tasks so that existing tests do not break. Fixes: https://tracker.ceph.com/issues/41666 Signed-off-by: Sridhar Seshasayee <[email protected]> (cherry picked from commit 33c647e8114b37404d8d62a08c85664cea709118) Conflicts: PendingReleaseNotes - Added release notes under 14.2.9 qa/standalone/mon/health-mute.sh - Deleted the script as 'health mute/unmute' cmd is unavailable in nautilus qa/tasks/ceph.conf.template - Removed a flag not available in nautilus src/common/options.cc - Removed a flag not available in nautilus src/osd/OSDMap.cc
nautilus: mgr/dashboard: Updated existing E2E tests to match new format Reviewed-by: Laura Paduano <[email protected]>
nautilus: mgr/telemetry: catch exception during requests.put Reviewed-by: Dan Mick <[email protected]>
nautilus: mgr/dashboard: Pool read/write OPS shows too many decimal places Reviewed-by: Stephan Müller <[email protected]> Reviewed-by: Laura Paduano <[email protected]>
nautilus: mgr/dashboard: do not show RGW API keys if only read-only privileges Reviewed-by: Lenz Grimmer <[email protected]> Reviewed-by: Laura Paduano <[email protected]> Reviewed-by: Ernesto Puerta <[email protected]> Reviewed-by: Volker Theile <[email protected]>
nautilus: pybind/mgr/*: fix config_notify handling of default values Reviewed-by: Sage Weil <[email protected]>
0b369e1 masked the original behaviour of '-o' which was to indicate 'outfile' as documented in the man page. Changing object-size to capital o will restore the original behaviour. Fixes: https://tracker.ceph.com/issues/42477 Signed-off-by: Brad Hubbard <[email protected]> (cherry picked from commit d2b806070de5f56474283035a7ffc91fe96afbab) Conflicts: PendingReleaseNotes: trivial resolution
nautilus: osd/PeeringState: transit async_recovery_targets back into acting before backfilling Reviewed-by: xie xingguo <[email protected]> Reviewed-by: Neha Ojha <[email protected]>
nautilus: mon/FSCommands: Fix 'add_data_pool' command and 'fs new' command Reviewed-by: Nathan Cutler <[email protected]>
…utilus nautilus: os/bluestore: open DB in read-only when expanding DB/WAL Reviewed-by: Adam Kupczyk <[email protected]>
nautilus: mon: Get session_map_lock before remove_session Reviewed-by: xie xingguo <[email protected]>
nautilus: os/bluestore: Don't pollute old journal when add new device Reviewed-by: Igor Fedotov <[email protected]>
nautilus: bluestore/bdev: initialize size when creating object Reviewed-by: Igor Fedotov <[email protected]>
nautilus: qa/distros: point {centos,rhel}_latest.yaml to 7.8
Reviewed-by: Yuri Weinstein <[email protected]>
Reviewed-by: David Galloway <[email protected]>
Reviewed-by: Laura Paduano <[email protected]> Reviewed-by: Tiago Melo <[email protected]>
nautilus: rgw: increase log level for same or older period pull msg Reviewed-by: Casey Bodley <[email protected]>
Signed-off-by: Patrick Donnelly <[email protected]> (cherry picked from commit 6ea2730ea669b6cf7ac668f606f63628ef8918b7) Conflicts: qa/tasks/cephfs/test_pool_perm.py - manually modify the py3 compatibility related fix
Fixes: https://tracker.ceph.com/issues/43515 Signed-off-by: Patrick Donnelly <[email protected]> (cherry picked from commit ae28ed11d7b9b1dac7dab1e8f5dc0c9643b5de76)
disable the TOO_FEW_PGS warning, as 1ac34a5ea3d1aca299b02e574b295dd4bf6167f4 is not backported to mimic, we will have TOO_FEW_PGS warnings when a healthy cluster is expected when upgrading from mimic. this change disables this warning by setting "mon_pg_warn_min_per_osd" to "0". this change is not cherry-picked from master. as 1ac34a5ea3d1aca299b02e574b295dd4bf6167f4 is already included by master, and we don't perform upgrade from mimic on master branch. Signed-off-by: Kefu Chai <[email protected]>
nautilus: mgr: force purge normal ceph entities from service map Reviewed-by: Ramana Raja <[email protected]>
nautilus: qa: use small default pg count for CephFS pools Reviewed-by: Ramana Raja <[email protected]>
nautilus: ceph-fuse: link to libfuse3 and pass "-o big_writes" to libfuse if libfuse < 3.0.0 Reviewed-by: Zheng Yan <[email protected]> Reviewed-by: Xiubo Li <[email protected]> Reviewed-by: Ramana Raja <[email protected]>
…pg-per-osd-warning nautilus: qa/suites/upgrade: disable min pg per osd warning Reviewed-by: Yuri Weinstein <[email protected]>
nautilus: rgw/notifications: backporting features and bug fix Reviewed-by: Casey Bodley <[email protected]>
nautilus: qa: fix various py3 cephfs qa bugs x2 Reviewed-by: Kefu Chai <[email protected]> Reviewed-by: Ramana Raja <[email protected]>
nautilus: mgr/volumes: Add interface to get subvolume metadata Reviewed-by: Ramana Raja <[email protected]>
nautilus: client: reset requested_max_size if file write is not wanted Reviewed-by: Ramana Raja <[email protected]> Reviewed-by: Greg Farnum <[email protected]> Reviewed-by: Yan, Zheng <[email protected]>
…enial nautilus: selinux: Allow getattr access to /proc/kcore Reviewed-by: Neha Ojha <[email protected]> Reviewed-by: Yuri Weinstein <[email protected]>
…s-wo-object-size nautilus: qa/suites/rados: use default objectsize for upgrade tests Reviewed-by: Yuri Weinstein <[email protected]> Reviewed-by: Brad Hubbard <[email protected]>
nautilus: qa/tasks/ceph.py: do not use option mimic does not understand Reviewed-by: Yuri Weinstein <[email protected]> Reviewed-by: Neha Ojha <[email protected]>
Fixes: https://tracker.ceph.com/issues/45515 Signed-off-by: Alfonso Martínez <[email protected]>
Reviewed-by: Alfonso Martínez <[email protected]> Reviewed-by: Ernesto Puerta <[email protected]> Reviewed-by: Laura Paduano <[email protected]>
…error Reviewed-by: Ernesto Puerta <[email protected]> Reviewed-by: Laura Paduano <[email protected]> Reviewed-by: Volker Theile <[email protected]>
nautilus: mgr/dashboard: fixing RBD purge error in backend Reviewed-by: Alfonso Martínez <[email protected]> Reviewed-by: Ernesto Puertat <[email protected]> Reviewed-by: Laura Paduano <[email protected]>
nautilus: monitoring: Fix pool capacity incorrect Reviewed-by: Alfonso Martínez <[email protected]> Reviewed-by: Patrick Seidensal <[email protected]> Reviewed-by: Laura Paduano <[email protected]>
nautilus: mgr/dashboard: fix backport #33764
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This PR aims to upgrade skyhook from luminous to nautilus