ceph学习笔记

ceph

ceph osd lspools


rbd ls -p testpool


#查看 ceph 集群中有多少个 pool,并且每个 pool 容量及利 用情况
rados df


ceph -s


ceph osd tree
ceph df

ceph versions

ceph osd pool ls

ceph osd crush rule dump

ceph auth print-key client.admin

ceph orch host ls

ceph crash ls

ceph osd pool stats

ceph df detail
ceph osd stat

ceph mon stat

查看image rbd
rbd ls -p kube

ceph osd df
ceph osd pool autoscale-status


ceph： 
10.240.62.11/12/13
root:autelceph2  


用户名：autel
密码：Autonomy@Autel

13 Autel#3913

[root@ceph-admin ~]# ceph mgr services
{
    "dashboard": "https://10.250.53.152:8443/",
    "prometheus": "http://10.250.53.152:9283/"
}



kubectl logs -f qinzhao-cache-resunet-demo-pipeline-wbkkh-2890309351 -n qinzhao -c lustre-importer-preload

kubectl get nodes "-o=custom-columns=NAME:.metadata.name,GPU:.status.allocatable.nvidia\.com/gpu"

apps/jupyter/jupyter-web-app/upstream/base/configs/spawner_ui_config.yaml

 kustomize build apps/jupyter/jupyter-web-app/upstream/overlays/istio | kubectl apply -f -

kustomize build  apps/tensorboard/tensorboard-controller/upstream/overlays/kubeflow | kubectl apply -f -

kubectl get pods -n kubeflow -l kustomize.component=profiles

https://www.amazonaws.cn/ec2/instance-types/

kubectl taint node autel-poweredge-r750 nodetype=T4:NoExecute
kubectl taint node autelrobotics-gpu10 nodetype=RTX3090:NoExecute
autelrobotics-gpu10
kubectl taint node autel-poweredge-r750 nodetype:NoExecute-

 kubectl taint node autelrobotics-gpu09 nodetype:NoSchedule-
 
  kubectl taint node autelrobotics-gpu09 nodetype:NoExecute-
  
  nodegroup=gpu:NoSchedule
  
  kubectl taint nodes autelrobotics-gpu02 nodegroup=gpu:NoSchedule
  
  kubectl label node autelrobotics-gpu02 gputype=A40
  
  lsof -n -P -i:22
  strace

kubectl get csinode
查看活跃进程个数
top -H -p 1

kubectl create secret tls ai-tls \
    --namespace ai-test \
    --key tls.key \
    --cert tls.pem

https://github.com/NVIDIA/nvidia-docker/issues/1678
nvidia-container-cli -k -d /dev/tty info


ls -l /dev/char
cat /etc/nvidia-container-runtime/config.toml

stat -fc %T /sys/fs/cgroup/

sar -n TCP,ETCP 1

 fdisk -l
 ldd
 
 # 修改后，重新挂载生效
# mount -o remount /dev/shm

nstat
mpstat -P all 1
slabtop
pcstat

netstat -ant | awk '{print $6}' | sort | uniq -c | sort -n

dmesg -T
pmap -x  1649 | sort -k 3 -n -r
cat /proc/1649/smaps | grep 7f4250021000
dump memory memory.dump 0x7f2340539000 0x7f235d553000

strings memory.dump

pidstat -p pid -r 1 1000
sudo ./stackcount ip_output

dmesg -Tw
perf
NetHogs

iftop -i eth0 -P -N
./opensnoop -Tn snmp-pass
slabtop

nfsstat -c

du -ah --max-depth=2 /var/log |sort -rh |head -10

./fileslower
ulimit -a


解决显存释放问题:
fuser -v /dev/nvidia*

 lsof -Pni
 
 netstat -n | awk '/^tcp/ {++S[$NF]} END {for(a in S) print a, S[a]}'
 

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159

NFS运维:

systemctl status rpcbind nfs-server
nfsiostat
dmesg | grep nfs
exportfs -v
mpstat -P ALL 1
ss -t -a |grep "IP"

nfsstat -c
iostat 

iostat -d -x -k 1

netstat -an | "IP:2049"

dstat
ps aux | grep /app

https://learnku.com/articles/39851
https://zhuanlan.zhihu.com/p/614314627

 fdisk -l
 blkid
 
 nfsiostat  1
 
 sar -b 1
 
 iostat -m -d /dev/md0 1
 
 strace -p pid 查看进程当前调用栈，查死循环或者卡顿时极为有用
strace -eopen /usr/local/kk-mail/service/dovecot/sbin/dovecot  查看进程当前打开了哪些文件 

cat /proc/715765/*/task/stack

/proc/12544/task/12873/stack
systemtap

cat /var/log/Xorg.0.log |grep -i "nvidia"

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39