Docs

Volcengine Kubernetes Engine

cluster management

How do I obtain node information of a cluster?

When an exception occurs in node resources in a VKE cluster, you can use the diagnosis feature provided by the AIOps Suite to locate issues in cluster nodes. For details, see Node diagnosis. If the node diagnosis feature does not meet your business requirements, you can use a script on the abnormal node to collect key node information. This facilitates subsequent diagnosis of node failures.

Prerequisites

To save node information in TOS, create a TOS bucket first. For details, see Creating buckets.

The TOS bucket needs to be in the same region as the cluster.

Procedure

Step 1: Obtain the information collection script

Use ECS Terminal to log in to the node. For details, see ECS Terminal.

Run the following command to download the information collection script.

curl -O https://vke-diagnostics-cn-beijing.tos-cn-beijing.volces.com/node_diagnostics.sh

tip

The command for obtaining the node information collection script varies based on different regions where your cluster is deployed and public network access statuses.

When your cluster is deployed in the China (Beijing) region, you can directly run the preceding command to obtain the script.

When your cluster is deployed in other regions:

If Public NAT for internet endpoint is enabled for your cluster, you can directly run the preceding command to obtain the script.

If Public NAT for internet endpoint is not enabled for your cluster, replace the public domain name (volces.com) in the preceding command with the private domain name (ivolces.com), and replace RegionID with the RegionID of your cluster. For example, when your cluster is deployed in the China (Shanghai) region, you need to modify the command to curl -O https://vke-diagnostics-cn-shanghai.tos-cn-shanghai.ivolces.com/node_diagnostics.sh. For more information about regions where VKE is available, see Regions and availability zones.

Run the following command to grant execution permissions to the script.

chmod +x node_diagnostics.sh

Step 2: Run the information collection script

(Optional) Run the following command to configure the modules for information collection. If no modules are configured, all information related to nodes will be collected.

bash node_diagnostics.sh --modules ${module-name-1},${module-name-2}

tip

You can run the bash node_diagnostics.sh --modules-list command to view all available modules. For example, to collect only network and memory information, configure the command bash node_diagnostics.sh --modules network,memory.

(Optional) Specify the location where the collected results are stored. To save collected results in TOS, run the following command to configure the AK/SK environment variables for your Volcano Engine account. If you do not specify a location, the collected information will be saved in the local path of the node.

export ACCESS_KEY_ID=${your_ak}
export ACCESS_KEY_SECRET=${your_sk}

The following table describes the parameters.

Parameter	Description
ACCESS_KEY_ID	The AK of your Volcano Engine root user. For instructions on obtaining the AK, see Access key.
ACCESS_KEY_SECRET	The SK of your Volcano Engine root user. For instructions on obtaining the SK, see Access key.

tip

Make sure that the account to which the AK/SK belongs has write permissions on the TOS bucket. Otherwise, the collected information will fail to be written. For TOS permission descriptions, see IAM policy overview.

Run the collection script.

Use the following command to run the script and save the result locally. The file path is /tmp/{hostname}_{timestamp}.

bash node_diagnostics.sh

Use the following command to run the script and save the result in the TOS bucket.

bash node_diagnostics.sh --bucket my-bucket --region cn-beijing

The following table describes the parameters.

Parameter
Description
--bucket
The TOS bucket name.
--region
The RegionID of the TOS bucket.

Validating results

After the script is run, the system displays the following example. The name of the generated log file is unique each time the diagnosis script is run. This topic uses iv-yeheeqetq8ug9cb7s1ki_20260309202204.tar.gz as an example. The actual result depends on the actual environment.

...
[Mon Mar 9 14:54:42 2026] intel_pstate: CPU model not supported
[Mon Mar 9 14:54:42 2026] ledtrig-cpu: registered to indicate activity on CPUs
[Mon Mar 9 14:54:42 2026] drop_monitor: Initializing network drop monitor service
[Mon Mar 9 14:54:42 2026] NET: Registered protocol family 10
[Mon Mar 9 14:54:42 2026] Segment Routing with IPv6
[Mon Mar 9 14:54:42 2026] mip6: Mobile IPv6
[Mon Mar 9 14:54:42 2026] NET: Registered protocol family 17
[Mon Mar 9 14:54:42 2026] mpls_gso: MPLS GSO support
[Mon Mar 9 14:54:42 2026] IPI shorthand broadcast: enabled
[Mon Mar 9 14:54:42 2026] sched_clock: Marking stable (1154488504, 106307620)->(1374330680, -113534556)
[Mon Mar 9 14:54:42 2026] registered taskstats version 1
[Mon Mar 9 14:54:42 2026] Loading compiled-in X.509 certificates
[Mon Mar 9 14:54:42 2026] zswap: loaded using pool lzo/zbud
[Mon Mar 9 14:54:42 2026] page_owner is disabled
...

Run the following command to access the node information folder and view the information collected from different modules.

cd iv-yeheeqetq8ug9cb7s1ki_20260309202204/
ls -l

The expected result is as follows:

total 60
drwxr-xr-x 2 root root 4096 Mar 9 20:22 cloud_init
drwxr-xr-x 2 root root 4096 Mar 9 20:22 disk
drwxr-xr-x 2 root root 4096 Mar 9 20:22 dmesg
drwxr-xr-x 2 root root 4096 Mar 9 20:22 gpu
-rw-r--r-- 1 root root 128 Mar 9 20:22 instance
drwxr-xr-x 5 root root 4096 Mar 9 20:22 kubernetes
drwxr-xr-x 2 root root 4096 Mar 9 20:22 memory
drwxr-xr-x 2 root root 4096 Mar 9 20:22 mgpu
drwxr-xr-x 2 root root 4096 Mar 9 20:22 network
drwxr-xr-x 3 root root 4096 Mar 9 20:22 process
drwxr-xr-x 2 root root 4096 Mar 9 20:22 service
drwxr-xr-x 2 root root 4096 Mar 9 20:22 syslog
drwxr-xr-x 2 root root 4096 Mar 9 20:22 system_info
drwxr-xr-x 2 root root 4096 Mar 9 20:22 throttling
drwxr-xr-x 2 root root 4096 Mar 9 20:22 top

View the detailed information collected by the specified module. This topic uses disk information as an example.

cat disk/df

The expected result is as follows:

-----------------run df -Th------------------
Filesystem Type Size Used Avail Use% Mounted on
udev devtmpfs 7.7G 0 7.7G 0% /dev
tmpfs tmpfs 1.6G 43M 1.5G 3% /run
/dev/vda2 ext4 40G 3.1G 35G 9% /
tmpfs tmpfs 7.7G 0 7.7G 0% /dev/shm
tmpfs tmpfs 5.0M 0 5.0M 0% /run/lock
tmpfs tmpfs 7.7G 0 7.7G 0% /sys/fs/cgroup
...

Last updated: 2026.07.23 14:31:47

Parameter	Description
--bucket	The TOS bucket name.
--region	The RegionID of the TOS bucket.