multipath devices typically show as 2 device nodes, so
gather-facts needs to use the serial number of the
device to ensure the disk/flash counts and capacity
calculations are not twice as big as they should be.
In addition to logic changes,
- the device data includes new fields: disk_type (hdd|flash),
serial, and alt_dev_name. The alt_dev_name shows the
alternate device name for the same serial number (when a
serial number is not present, this is just '')
- enclosure information (SES only) is included based
on device serial number. This maps serial numbers to slots
and provides a breakdown of the slot states
- the supported block device list now excludes loopback and
mdraid (/dev/md*) devices
Fixes: https://tracker.ceph.com/issues/55489
Signed-off-by: Paul Cuzner <pcuzner@redhat.com>
in a recent change in Seastar, DERIVE metric was dropped in favor
of COUNTER.
in this change, all DERIVE metrics are replaced with COUNTER metrics.
Signed-off-by: Kefu Chai <tchaikov@gmail.com>
don't rely on the ceph manager task to parse a config file. each rgw
could be using a different config. instead, revert to an s3tests
override called 'with-sse-s3'
this way, the only job that enables sse-s3, vault_transit.yaml, contains
both the 'rgw crypt sse s3' configurables, and the flag to enable the
associated test cases
Signed-off-by: Casey Bodley <cbodley@redhat.com>
Description:
- While using `get` command, `local_path` parameter is optional. Changing it
to mandatory.
- Rationale: Till now, there used to be a default path of `local_path` as
`default='.'` but wasn't mentioned anywhere. It led to confusion.
On top of it, considering get command to be a ssh inspired utlity,
or any other CLI tool that copies file between filesystems, source
and destination path are always mandatory. Therefore in order to
simulate this behavior in cephfs-shell`s command(s), my opinion is
to make get command accept both the paths.
- Added checks to make sure:
1) File does exist at `remote_path`
2) File with the same name doesn't exist in `local_path`
3) Removed code that would run through the directory and if it finds
nothing in `root_src_dir`, then it will try to do:
`os.makedirs(root_dst_dir + b'/' + root_src_dir)`, but it will
never be empty as 1) takes care of it.
Fixes: https://tracker.ceph.com/issues/55216
Signed-off-by: dparmar18 <dparmar@redhat.com>
Descrption: When using the get command on a single file, it would append the remote path to
source path and create directories that shouldn't be created. For instance,
file 'foo.txt' resides at `/dir1/dir2/` and `get` command is used to copy it to
/tmp/foo then it would do `/tmp/dir1/dir2/foo.txt` which is not the expected
behavior. Therefore this PR intends to correct this behavior.
Fixes: https://tracker.ceph.com/issues/55112
Signed-off-by: Dhairya Parmar <dparmar@redhat.com>
Rationale: get and put now demand both the paths mandatorily.
Also testing of get and put without target paths
have been take of in other tests in class TestGetAndPut().
Signed-off-by: Dhairya Parmar <dparmar@redhat.com>
Description:
- `put` command didn't display any error when file at local_path was not
found. This PR intends to add that check.
- Rationale: Till now, there used to be a default path of `remote_path` as
`default='.'` but wasn't mentioned anywhere. It could lead to confusion.
On top of it, considering put command to be a ssh inspired utlity,
or any other CLI tool that copies file between filesystems, source
and destination path are always mandatory. Therefore in order to
simulate this behavior in cephfs-shell`s command(s), my opinion is
to make put command accept both the paths.
Fixes: https://tracker.ceph.com/issues/55242
Signed-off-by: dparmar18 <dparmar@redhat.com>
Prometheus is currently using only the hostname in the 'generatorURL' of an alert which causes issues when clicking on the URL in the Ceph Dashboard or somewhere else, because in most cases the hostname of the node that is running the Prometheus container is not resolvable.
To fix that the command line argument '--web.external-url' must be appended in the systemd unit file of the Prometheus container, e.g. '--web.external-url http://foo.bar:9095' whereas a FQDN hostname is used.
Fixes: https://tracker.ceph.com/issues/55595
Signed-off-by: Volker Theile <vtheile@suse.com>
Fix the confusing combination of default-member-initializers
for some members and direct initialization of others in
the ctor.
Signed-off-by: Ronen Friedman <rfriedma@redhat.com>
before lua script is being executed, we keep the tracer runtime configuration value, and then decides whether to trace or not the request based on the value that maybe changed during lua exeuction, so we can disable/enable tracing for request even if the tracer is in the opposite state at the same time
Signed-off-by: Omri Zeneva <ozeneva@redhat.com>
introducing the scrubber_generators to create
scrubber test data, and the scrubber_test_datasets
for pre-prepared test configurations.
Signed-off-by: Ronen Friedman <rfriedma@redhat.com>
