summaryrefslogtreecommitdiffstats
path: root/src/cephadm (follow)
Commit message (Collapse)AuthorAgeFilesLines
* mgr/cephadm: ceph nvmeof monitor supportAlexander Indenbaum2024-07-311-3/+14
| | | | Signed-off-by: Alexander Indenbaum <aindenba@redhat.com>
* Merge pull request #57712 from rhcs-dashboard/dashboard-nvmf-mtls-confNizamudeen A2024-07-121-0/+18
|\ | | | | | | | | | | | | | | mgr/dashboard: use secure_channel for grpc requests Reviewed-by: Adam King <adking@redhat.com> Reviewed-by: baum <NOT@FOUND> Reviewed-by: Ernesto Puerta <epuertat@redhat.com> Reviewed-by: Redouane Kachach <rkachach@redhat.com>
| * python-common/service_spec: add root_ca_cert to nvmeof specAdam King2024-07-081-1/+7
| | | | | | | | | | | | | | | | Also improves the error messaging around when spec/key attributes are missing when enable_auth is set to true Signed-off-by: Adam King <adking@redhat.com> (cherry picked from commit 9212914be65fe3adde2108f5a2cfd2587d17c0ff)
| * mgr/cephadm: allow passing client/server cert/key in nvmeof specAdam King2024-07-081-0/+12
| | | | | | | | | | | | | | | | | | | | | | | | | | Before this patch the client/server cert/key fields were just filepaths that told the nvmeof gw daemon where to look for the cert/key. There's not much reason why users would care where in the nvmeof gw container the cert goes. It's more useful to use these fields as a way to pass the certs/keys to the daemon and then just hardcode where in the container we'll place the certs/keys Signed-off-by: Adam King <adking@redhat.com> (cherry picked from commit e9fca39092348e6c08022341116875e831c175f0)
* | Merge pull request #58321 from adk3798/jinja2-cephadm-rhel9Adam King2024-07-101-0/+27
|\ \ | | | | | | | | | | | | cephadm: make cephadm compatible with jinja2 2.11.3 Reviewed-by: Adam King <adking@redhat.com>
| * | cephadm: make cephadm compatible with jinja2 2.11.3Adam King2024-07-031-0/+27
| |/ | | | | | | | | | | | | That version is what's currently being shipped in RPMs on rhel9 Signed-off-by: Adam King <adking@redhat.com>
* | mgr/cephadm: introducing nobody/nogroup constantsRedouane Kachach2024-07-095-6/+12
| | | | | | | | Signed-off-by: Redouane Kachach <rkachach@ibm.com>
* | mgr/cephadm: adding new cephadm service mgmt-gatewayRedouane Kachach2024-07-095-1/+199
|/ | | | | | | | | | | | | adding mgmt-gateway, a new cephadm service based on nginx, to act as the front-end and single entry point to the cluster. This gateway offers unified access to all Ceph applications, including the Ceph dashboard and monitoring tools (Prometheus, Grafana, ..), while enhancing security and simplifying access management through nginx. Fixes: https://tracker.ceph.com/issues/66095 Signed-off-by: Redouane Kachach <rkachach@ibm.com>
* Merge pull request #57975 from dvanders/dvanders_ipv4Adam King2024-07-011-0/+6
|\ | | | | | | | | cephadm: disable ms_bind_ipv4 if we will enable ms_bind_ipv6 Reviewed-by: Adam King <adking@redhat.com>
| * cephadm: disable ms_bind_ipv4 if we will enable ms_bind_ipv6Dan van der Ster2024-06-111-0/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | While bootstrapping an ipv6 cluster with an ipv6 initial mon, cephadm correctly enables ms_bind_ipv6=true. However it leaves ms_bind_ipv4 as it's default (true). As a result, daemons (osd, mds, ...) will attempt to bind to both ipv6 and ipv4. Usually this results in an osdmap and fsmap like the following: ``` osd.2 up in weight 1 up_from 26 up_thru 909 down_at 0 last_clean_interval [0,0) [v2:[xxxx:4f8:d0:4401:3::29]:6800/3680761436,v1:[xxxx:4f8:d0:4401:3::29]:6801/3680761436,v2:0.0.0.0:6802/3680761436,v1:0.0.0.0:6803/3680761436] [v2:[xxxx:4f8:d0:4401:3::29]:6804/3680761436,v1:[xxxx:4f8:d0:4401:3::29]:6805/3680761436,v2:0.0.0.0:6806/3680761436,v1:0.0.0.0:6807/3680761436] exists,up 0978a571-cd00-4eba-b00b-f863603a9a70 ``` ``` [mds.cephfs.ceph-test-3.isityv{-1:793154} state up:standby seq 1 join_fscid=1 addr [v2:[xxxx:4f8:d0:4401:3::29]:6832/2213688825,v1:[xxxx:4f8:d0:4401:3::29]:6833/2213688825,v2:0.0.0.0:6834/2213688825,v1:0.0.0.0:6835/2213688825] compat {c=[1],r=[1],i=[7ff]}] ``` Dual stack is not support by kernels (https://tracker.ceph.com/issues/49581) which leads to hard to debug issues for the end users. (corrupt map messages in dmesg). Fix by disabling ms_bind_ipv4 in the case ipv6 is desired. Fixes: https://tracker.ceph.com/issues/66436 Signed-off-by: Dan van der Ster <dan.vanderster@clyso.com> Signed-off-by: Joshua Blanch <joshua.blanch@clyso.com>
* | Merge pull request #57955 from phlogistonjohn/jjm-fix-issue66389-apparmorAdam King2024-06-173-76/+120
|\ \ | | | | | | | | | | | | cephadm: fix apparmor profiles with spaces in the names Reviewed-by: Adam King <adking@redhat.com>
| * | cephadm: rename test_enclosure to test_host_factsJohn Mulligan2024-06-111-0/+0
| | | | | | | | | | | | | | | | | | | | | | | | There was a whole file dedicated to the enclosure class from host_facts, but no other tests for host facts. Rename the enclosure test file to cover all of host_facts module (for the future). Signed-off-by: John Mulligan <jmulligan@redhat.com>
| * | cephadm: black-format the file test_enclosure.pyJohn Mulligan2024-06-111-15/+24
| | | | | | | | | | | | | | | | | | | | | A bunch of stray whitespace in this file called out to be cleaned up. Black did that job and fixed up some other formatting stuff. Signed-off-by: John Mulligan <jmulligan@redhat.com>
| * | cephadm: update hosts_facts to read apparmor profile names with spacesJohn Mulligan2024-06-112-7/+7
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Fixes: https://tracker.ceph.com/issues/66389 Update the host_facts class kernel_security method to correctly read apparmor profile names that have spaces in them. Update the test to verify this functionality. Original-version-by: Sebastian Marsching <sebastian.marsching-git-2016@aquenos.com> Signed-off-by: John Mulligan <jmulligan@redhat.com>
| * | cephadm: add a test case to cover reading apparmor profilesJohn Mulligan2024-06-111-1/+36
| |/ | | | | | | | | | | | | | | | | Add a test case that covers the HostFacts functionality regarding the apparmor kernel security (lsm) feature. Put it in the test_enclosure.py file for now because enclosure is part of host_facts module. Signed-off-by: John Mulligan <jmulligan@redhat.com>
* | src/cephadm/box: remove unused importsrkhudov2024-06-144-9/+2
| | | | | | | | Signed-off-by: rkhudov <rhudov@gmail.com>
* | Merge pull request #57601 from guits/cephadm-hostspec-fixesAdam King2024-06-122-84/+83
|\ \ | |/ |/| | | | | cephadm: _extract_host_info_from_*() refactor Reviewed-by: John Mulligan <jmulligan@redhat.com>
| * cephadm: _extract_host_info_from_*() refactorGuillaume Abrioux2024-05-272-84/+83
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The current implementation doesn't take into account the format is Yaml This can lead to issue when oob details are passed to host service spec. For instance, with the following host spec: ``` --- service_type: host addr: 1.2.3.4 hostname: node1 oob: username: root password: passw0rd addr: 127.0.0.1 ``` it is converted to a list like the following: ``` ['service_type: host', 'addr: 1.2.3.4', 'hostname: node1', 'oob:', 'username: root', 'password: passw0rd', 'addr: 127.0.0.1'] ``` It was (probably) assumed that the pattern `addr:` would be present only once. With the introduction of node-proxy, this isn't true anymore. Now that the cephadm binary can embed some external libraries we can leverage pyyaml. The idea is to use proper yaml format instead so it is easier to process the data. Fixes: https://tracker.ceph.com/issues/66165 Signed-off-by: Guillaume Abrioux <gabrioux@ibm.com>
* | Merge pull request #57829 from gukaifeng/adjust_bootstrap_config_prioritiesAdam King2024-06-111-17/+20
|\ \ | | | | | | | | | | | | cephadm: the user config is preferred during the bootstrap Reviewed-by: Adam King <adking@redhat.com>
| * | cephadm: the user config is preferred during the bootstrapgukaifeng2024-06-051-17/+20
| |/ | | | | | | Signed-off-by: gukaifeng <gukaifeng@xiaomi.com>
* | Merge pull request #57685 from tchaikov/replace-pkg_resources-with-importlibAdam King2024-06-111-14/+23
|\ \ | | | | | | | | | | | | | | | cephadm: use importlib.metadata for querying ceph_iscsi's version Reviewed-by: Adam King <adking@redhat.com> Reviewed-by: John Mulligan <jmulligan@redhat.com>
| * | cephadm: use importlib.metadata for querying ceph_iscsi's versionKefu Chai2024-05-261-0/+5
| | | | | | | | | | | | | | | | | | | | | | | | | | | use importlib.metadata for querying ceph_iscsi's version and fallback to pkg_resources. as the former is only available in Python 3.8, while the latter is deprecated. Refs https://tracker.ceph.com/issues/66201 Signed-off-by: Kefu Chai <tchaikov@gmail.com>
| * | cephadm: extract python() helper to execute python statementKefu Chai2024-05-261-14/+18
| |/ | | | | | | | | | | | | | | | | to prepare for a change to use importlib, then fallback to pkg_resources. as the former is only available in Python 3.8, while the latter is deprecated. Refs https://tracker.ceph.com/issues/66201 Signed-off-by: Kefu Chai <tchaikov@gmail.com>
* | Merge pull request #57389 from thmour/patch-1Adam King2024-06-111-2/+2
|\ \ | |/ |/| | | | | cephadm: CephExporter doesn't bind to IPv6 in dual stack Reviewed-by: Adam King <adking@redhat.com>
| * Fix CephExporter protocol bind logicMouratidis Theofilos2024-05-101-2/+2
| | | | | | | | | | In a dual stack configuration ceph-exporter binds to ipv4 only and the metrics fail in ipv6 Signed-off-by: Mouratidis Theofilos <mtheofilos@gmail.com>
* | cephadm: change loki/promtail default image tagsGuillaume Abrioux2024-05-131-2/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This commit changes the default images for both loki and promtail containers. Also, to allow this update we need to update the configuration of loki in order to add a new storage schema configuration: ``` - from: 2024-05-03 store: tsdb object_store: filesystem schema: v13 index: prefix: index_ period: 24h ``` This is because 3.0.0 uses new structured metadata. See [1][2]. [1] https://grafana.com/docs/loki/latest/operations/storage/schema/ [2] https://grafana.com/docs/loki/latest/get-started/labels/structured-metadata/#when-to-use-structured-metadata Fixes: https://tracker.ceph.com/issues/65784 Signed-off-by: Guillaume Abrioux <gabrioux@ibm.com>
* | cephadm: Adding support to pass --no-cgroups-split flag when adopting legacy ↵Gilad Sid2024-05-051-0/+5
|/ | | | | | daemons Signed-off-by: Gilad Sid <sid.gilad@gmail.com>
* Merge pull request #56961 from adk3798/agent-check-error-before-jsonAdam King2024-04-301-2/+2
|\ | | | | | | | | cephadm: have agent check for errors before json loading mgr response Reviewed-by: <mfritch@suse.com>
| * cephadm: have agent check for errors before json loading mgr responseAdam King2024-04-221-2/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Currently, since it tries to json.loads the response payload before checking the return code, if there was an error it fails with Failed to send metadata to mgr: the JSON object must be str, bytes or bytearray, not ConnectionRefusedError which is masking the actual failure. Also adds more context to the RuntimeError raised Fixes: https://tracker.ceph.com/issues/65553 Signed-off-by: Adam King <adking@redhat.com>
* | Merge pull request #56350 from phlogistonjohn/jjm-mgr-smb-moduleAdam King2024-04-301-2/+11
|\ \ | | | | | | | | | | | | mgr: add a new smb cluster & share module Reviewed-by: Adam King <adking@redhat.com>
| * | cephadm: handle user_sources uri values in smb daemonJohn Mulligan2024-04-261-2/+11
| | | | | | | | | | | | | | | | | | | | | | | | | | | When a smb daemon is being configured it may have user_sources - a field containing uris that are supplemental configurations expected to define users and/or groups for a non-AD member server. Ensure these uris get passed to the env var for the config uris to get processed. Signed-off-by: John Mulligan <jmulligan@redhat.com>
* | | Merge pull request #56817 from guits/cephadm-check-apply-spec-fileAdam King2024-04-301-3/+14
|\ \ \ | | | | | | | | | | | | | | | | | | | | cephadm: check if file exists when passing `--apply_spec` Reviewed-by: Adam King <adking@redhat.com> Reviewed-by: John Mulligan <jmulligan@redhat.com>
| * | | cephadm: check if file exists when passing `--apply_spec`Guillaume Abrioux2024-04-101-3/+14
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | cephadm deploys the cluster, fails and does a rollback. If the passed file doesn't exist we can make the CLI fail early instead. ``` ... omitted output ... Applying ../host-spec.yaml to cluster FileNotFoundError: [Errno 2] No such file or directory: '../host-spec.yaml' *************** Cephadm hit an issue during cluster installation. Current cluster files will be deleted automatically. To disable this behaviour you can pass the --no-cleanup-on-failure flag. In case of any previous broken installation, users must use the following command to completely delete the broken cluster: > cephadm rm-cluster --force --zap-osds --fsid <fsid> for more information please refer to https://docs.ceph.com/en/latest/cephadm/operations/#purging-a-cluster *************** Deleting cluster with fsid: 6e6a2dbe-f73a-11ee-8262-98be948800fd Traceback (most recent call last): File "/usr/lib64/python3.9/runpy.py", line 197, in _run_module_as_main return _run_code(code, main_globals, None, File "/usr/lib64/python3.9/runpy.py", line 87, in _run_code exec(code, run_globals) File "/tmp/tmpive4g9gs.cephadm.build/app/__main__.py", line 5615, in <module> File "/tmp/tmpive4g9gs.cephadm.build/app/__main__.py", line 5603, in main File "/tmp/tmpive4g9gs.cephadm.build/app/__main__.py", line 2693, in _rollback File "/tmp/tmpive4g9gs.cephadm.build/app/__main__.py", line 445, in _default_image File "/tmp/tmpive4g9gs.cephadm.build/app/__main__.py", line 2958, in command_bootstrap FileNotFoundError: [Errno 2] No such file or directory: '../host-spec.yaml' ``` Signed-off-by: Guillaume Abrioux <gabrioux@ibm.com>
* | | | Merge pull request #56487 from adk3798/bootstrap-custom-alertsAdam King2024-04-301-0/+13
|\ \ \ \ | |_|/ / |/| | | | | | | | | | | cephadm: allow passing custom prometheus alerts to bootstrap Reviewed-by: John Mulligan <jmulligan@redhat.com>
| * | | cephadm: allow passing custom prometheus alerts to bootstrapAdam King2024-03-261-0/+13
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This can't be done with the --config flag as the thing that needs to be changed is in the config-key store rather than being a config option. I don't think we have any option other than a new flag here. Makes use of the `ceph orch prometheus set-custom-alerts` command being added in another commit Signed-off-by: Adam King <adking@redhat.com>
* | | | Merge pull request #56942 from Svelar/fix_tox_cephadmRongqi Sun2024-04-251-1/+1
|\ \ \ \ | |_|_|/ |/| | | test/cephadm: extend timeout again
| * | | test/cephadm: extend timeout limit again to fix timeout issueRongqi Sun2024-04-231-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When run make check(arm64) on omani series servers, Shows: python: timeout after 60 seconds INFO root:call_wrappers.py:284 Non-zero exit code 124 from /home/jenkins-build/build/workspace/ceph-pull-requests-arm64/src/cephadm/.tox/py3/bin/python -c for i in range(1000000): print(i, flush=True) Seems like 60s is not enough, extend to 300s would be fine. Signed-off-by: Rongqi Sun <sunrongqi@huawei.com>
* | | | Merge pull request #56821 from phlogistonjohn/jjm-tox-no-basepythonAdam King2024-04-151-2/+0
|\ \ \ \ | | | | | | | | | | | | | | | | | | | | python: remove some vestigial tox basepython directives Reviewed-by: Adam King <adking@redhat.com>
| * | | | cephadm: remove vestigial basepython directivesJohn Mulligan2024-04-101-2/+0
| | |_|/ | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Remove what I think are vestigial configuration parameters setting `basepython` in tox.ini. These may have been useful during the change from python 2 to 3, but I don't see them serving any useful purpose at this current time. In fact they now interfere with the ability to globally override the basepython version of all testenv's at once. So let's just remove them. We can always add it back if we find an issue in the future (and document it!). Signed-off-by: John Mulligan <jmulligan@redhat.com>
* | | | Merge pull request #56609 from Svelar/fix_tox_cephadmRongqi Sun2024-04-151-1/+1
|\ \ \ \ | | |/ / | |/| | test/cephadm: fix timeout issue
| * | | test/cephadm: extend timeout limit to fix timeout issueRongqi Sun2024-04-111-1/+1
| | |/ | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When run make check(arm64) on omani series servers, eg. https://jenkins.ceph.com/job/ceph-pull-requests-arm64/54751/. Shows: python: timeout after 30 seconds INFO root:call_wrappers.py:284 Non-zero exit code 124 from /home/jenkins-build/build/workspace/ceph-pull-requests-arm64/src/cephadm/.tox/py3/bin/python -c for i in range(1000000): print(i, flush=True) Fixes: https://tracker.ceph.com/issues/65355 Signed-off-by: Rongqi Sun <sunrongqi@huawei.com>
* | | Merge pull request #56561 from phlogistonjohn/jjm-issue65122-maint-cmdAdam King2024-04-102-5/+15
|\ \ \ | | | | | | | | | | | | | | | | cephadm: fix host-maintenance command always exiting with a failure Reviewed-by: Adam King <adking@redhat.com>
| * | | cephadm: fix host-maintenance command always exiting with a failureJohn Mulligan2024-03-282-5/+15
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The host-maintenance command would always fail because command_maintenance always returns a string. This string is passed to sys.exit and thus always gets printed and causes a non-zero exit code. Fix the command line behavior by renaming the original function and adding a new command_maintenance that prints the string and returns an int like other command_* functions do. Fixes: https://tracker.ceph.com/issues/65122 Signed-off-by: John Mulligan <jmulligan@redhat.com>
* | | | Merge pull request #56481 from adk3798/test-cephadm-idmap-confAdam King2024-04-101-1/+2
|\ \ \ \ | |_|_|/ |/| | | | | | | | | | | cephadm: add idmap.conf to nfs sample file Reviewed-by: John Mulligan <jmulligan@redhat.com>
| * | | cephadm: add idmap.conf to nfs sample fileAdam King2024-03-261-1/+2
| | |/ | |/| | | | | | | | | | | | | | | | | | | | | | Otherwise deployment using this sample will fail with "ERROR: required file missing from config-json: idmap.conf" Fixes: https://tracker.ceph.com/issues/65155 Signed-off-by: Adam King <adking@redhat.com>
* | | Merge pull request #56641 from rhcs-dashboard/wip-65268-mainErnesto Puerta2024-04-091-1/+1
|\ \ \ | | | | | | | | | | | | | | | | mgr/dashboard: update NVMeoF gRPC: "listener add" changes Reviewed-by: gbregman <NOT@FOUND>
| * | | mgr/dashboard: update NVMeoF gRPCErnesto Puerta2024-04-081-1/+1
| | |/ | |/| | | | | | | | | | | | | | | | To include latest changes to the listener add call. Fixes: https://tracker.ceph.com/issues/65268 Signed-off-by: Ernesto Puerta <epuertat@redhat.com>
* / | cephadm: pin pyfakefs version for tox testsAdam King2024-04-081-1/+1
|/ / | | | | | | | | | | | | | | 5.4.0 seems to have caused a number of permission issues in the test. In order to unblock make check, let's pin it to 5.3.5 for now. Signed-off-by: Adam King <adking@redhat.com>
* | Merge pull request #56304 from phlogistonjohn/jjm-tox-default-envsAdam King2024-03-272-10/+3
|\ \ | |/ |/| | | | | python: allow unit tests to use tox default envs Reviewed-by: Adam King <adking@redhat.com>
| * cephadm: add format-black reminder commentJohn Mulligan2024-03-201-0/+2
| | | | | | | | | | | | | | | | | | Add a comment to tox.ini for people who might not know about the relatively new `check-black` and `format-black` envs. While check-black is part of the default envlist (as it is read-only) someone may not know about the read-write version `format-black`. Signed-off-by: John Mulligan <jmulligan@redhat.com>
generated by cgit v1.2.3 (git 2.39.1) at 2025-01-19 17:19:50 +0100