Hi xuchenhuig
> "client_metadata": {
> "client_features": {
> "feature_bits": "0x00000000000000ff"
> },
From the above feature_bits, it says you are still using the old
kernel, which hasn't support the io read/write metrics yet.
Please make sure the kernel you are using has downstreamed at least the
following commit:
https://github.com/torvalds/linux/commit/903f4fec78dd05a48fdccdf4539c040fb2d5bbf4
BRs
-- Xiubo
On 3/19/22 10:47 AM, xuchenhuig wrote:
Hi,
I haven`t seen my mail in
https://lists.ceph.io/hyperkitty/list/ceph-users@xxxxxxx/.
<https://lists.ceph.io/hyperkitty/list/ceph-users@xxxxxxx/.> I don`t
knon why. I find PR in https://github.com/ceph/ceph/pull/43784
<https://github.com/ceph/ceph/pull/43784> about you. Can you help me
resolve my question ?
Thanks
xuchenhuig
xuchenhuig@xxxxxxxxx
<https://maas.mail.163.com/dashi-web-extend/html/proSignature.html?ftlId=1&name=xuchenhuig&uid=xuchenhuig%40gmail.com&iconUrl=https%3A%2F%2Fmail-online.nosdn.127.net%2Fsmdc683e2de3c2d8b09490ad0fc174868b.jpg&items=%5B%22xuchenhuig%40gmail.com%22%5D>
签名由 网易邮箱大师 <https://mail.163.com/dashi/dlpro.html?from=mail81>
定制
--------- Forwarded Message ---------
From： 许晨辉 <xuchenhuig@xxxxxxxxx> <mailto:xuchenhuig@xxxxxxxxx>
Date： 03/18/2022 19:27
To： ceph-users@xxxxxxx <ceph-users@xxxxxxx> <mailto:ceph-users@xxxxxxx>
Subject： Cephfs:Can`t get read/write io size metrics by kernel client
Hello,
Question:I want to monitor cephfs client by get client session`s
read_io_sizes and write_io_sizes , but I can`t get read/write io size
metrics by kernel client when I exec 'ceph fs perf stats'. However,
ceph-fuse is ok.
Reproduce:
when I am copying 1GB data to cephfs, the cephfs perf stats show below:
# ceph fs perf stats
{"version": 1, "global_counters": ["cap_hit", "read_latency",
"write_latency", "metadata_latency", "dentry_lease", "opened_files",
"pinned_icaps", "opened_inodes", "read_io_sizes", "write_io_sizes"],
"counters": [], "client_metadata": {"client.499442": {"IP":
"v1:10.204.0.118", "hostname": "wc-testk8s-0-118", "root": "/",
"mount_point": "N/A"}}, "global_metrics": {"client.499442": [[0, 0],
[0, 0], [0, 0], [0, 0], [0, 0], [0, 0], [0, 0], [0, 0], [0, 0], [0,
0]]}, "metrics": {"delayed_ranks": [], "mds.0": {"client.499442": []}}}
All metrics is 0, seem to not support.
My Environment:
ceph version: v16.2.7
ceph kernel client：
release :CentOS Linux release 8.4.2105
kernel version: 4.14.105-19-0021
mount command:mount -t ceph
10.204.0.161:6789,10.204.0.148:6789,10.204.0.164:6789:/
/tmp/mount_dir/ -o
name=admin,secret=AQCfMTBiQtbaLBAANMTNA1Yc28Gc2H/dfvR4Gg==
client session info：
{
"id": 499442,
"entity": {
"name": {
"type": "client",
"num": 499442
},
"addr": {
"type": "v1",
"addr": "10.204.0.118:0",
"nonce": 2784782672
}
},
"state": "open",
"num_leases": 0,
"num_caps": 5,
"request_load_avg": 0,
"uptime": 524.59935749399995,
"requests_in_flight": 0,
"num_completed_requests": 0,
"num_completed_flushes": 0,
"reconnecting": false,
"recall_caps": {
"value": 0,
"halflife": 60
},
"release_caps": {
"value": 0,
"halflife": 60
},
"recall_caps_throttle": {
"value": 0,
"halflife": 1.5
},
"recall_caps_throttle2o": {
"value": 0,
"halflife": 0.5
},
"session_cache_liveness": {
"value": 2.462407850585298,
"halflife": 300
},
"cap_acquisition": {
"value": 0,
"halflife": 10
},
"delegated_inos": [],
"inst": "client.499442 v1:10.204.0.118:0/2784782672",
"completed_requests": [],
"prealloc_inos": [],
"client_metadata": {
"client_features": {
"feature_bits": "0x00000000000000ff"
},
"metric_spec": {
"metric_flags": {
"feature_bits": "0x"
}
},
"entity_id": "admin",
"hostname": "wc-testk8s-0-118",
"kernel_version": "4.14.105-19-0021",
"root": "/"
}
}
I have read the cephfs source code, but no result. I have a doubt
about parameter client_metadata.metric_spec.metric_flags.feature_bits.
Is this control valid_metrics? or My kernel version not support?
How do i get read/write io size metrics by kernel client?
Thanks.
*big xu
*邮箱：xuchenhuig@xxxxxxxxx
<https://maas.mail.163.com/dashi-web-extend/html/proSignature.html?ftlId=1&name=big+xu&uid=xuchenhuig%40gmail.com&iconUrl=https%3A%2F%2Fmail-online.nosdn.127.net%2Fqiyelogo%2FdefaultAvatar.png&items=%5B%22%E9%82%AE%E7%AE%B1%EF%BC%9Axuchenhuig%40gmail.com%22%5D>签名由
网易邮箱大师 <https://mail.163.com/dashi/dlpro.html?from=mail88> 定制
_______________________________________________
ceph-users mailing list -- ceph-users@xxxxxxx
To unsubscribe send an email to ceph-users-leave@xxxxxxx
