Procedure for collecting VSP One SDS Block dump log files

You can collect the dump log file of VSP One SDS Block.

This section describes how to collect dump log files by using the REST API/CLI. For details about how to collect dump log files by using the VSP One SDS Block Administrator, see Operation with dump log files in the VSP One SDS Block and SDS Cloud VSP One SDS Administrator Reference.

Note:

(Bare metal) If a failure occurs when you configure a storage cluster, add storage nodes, or replace storage nodes and you need to collect dump log files, you might need to collect them via the console interface rather than using the procedure described in this section. The following table shows cases requiring collection of dump log files by using the procedure described in this section, and cases requiring collection of dump log files by using the procedure described in Procedure for collecting dump log files via the console interface (Bare metal).

Operation for which a failure occurred	Storage nodes with which a storage cluster was being configured¹	Storage nodes being added	Storage nodes being replaced	Spare nodes	Other existing storage nodes
Configuring a storage cluster	Procedure in Procedure for collecting dump log files via the console interface (Bare metal)	-	-	-	-
Adding storage nodes	-	Procedure in Procedure for collecting dump log files via the console interface (Bare metal)²	-	-	Procedure in this section
Replacing storage nodes	-	-	Procedure in Procedure for collecting dump log files via the console interface (Bare metal)²	-	Procedure in this section
Other than above	-	-	-	Procedure in Procedure for collecting dump log files via the console interface (Bare metal)³	Procedure in this section
1. There might be a case where you only have to collect dump log files for only specific storage nodes. When you collect dump log files as described in Action to be taken for a failure that occurred during setup of each storage node (Bare metal), you only have to collect dump log files for the storage node on which the failure occurred. 2. The procedure in this section applies depending on the progress of the processing for the target storage node. The procedure to be used depends on whether Login:setup is displayed on the first line of the console interface login screen. You can choose the appropriate procedure according to the Procedure for collecting dump log files via the console interface (Bare metal). 3. Perform this procedure only when instructed to do so in the document or by customer support.

Note:

(Cloud for AWS) If you are collecting VSP One SDS Block dump log files to troubleshoot an error that occurred during storage cluster configuration, there are cases where you must follow the procedure as described in Procedure for collecting dump log files of VSP One SDS Block via Amazon S3 (Cloud for AWS) instead of the procedure in this section. In the AWS Management Console, verify the information for the root stack among the stacks you created during the storage cluster configuration. If the value of "InstallationDumpLogFile" is "S3 URI", follow the procedure as described in Procedure for collecting dump log files of VSP One SDS Block via Amazon S3 (Cloud for AWS). Otherwise, perform the procedure in this section.

(Cloud for Google Cloud) If you are collecting VSP One SDS Block dump log files to troubleshoot an error that occurred during storage cluster configuration, there are cases where you must follow the procedure as described in Procedure for collecting dump log files of VSP One SDS Block via Google Cloud Storage (Cloud for Google Cloud) instead of the procedure in this section. If a file named "Failed" exists in the directory specified for the "gsUri" parameter in the terraform.tfvars file for storage cluster configuration, perform the procedure described in Procedure for collecting dump log files of VSP One SDS Block via Google Cloud Storage (Cloud for Google Cloud). Otherwise, perform the procedure in this section.

(Cloud for Microsoft Azure) If you are collecting VSP One SDS Block dump log files to troubleshoot an error that occurred during storage cluster configuration, there are cases where you must follow the procedure as described in Procedure for collecting dump log files of VSP One SDS Block via Azure Blob Storage (Cloud for Microsoft Azure) instead of the procedure in this section. If the output item "InstallationStatus" of the deployment named "InstallationStatus" for the resource group (created for storage cluster configuration) is "Failed", perform the procedure described in Procedure for collecting dump log files of VSP One SDS Block via Azure Blob Storage (Cloud for Microsoft Azure). Otherwise, perform the procedure in this section.

Dump log files are created in the following file name:

When splitting of dump log files is not specified:

hsds_log_<YYYYMMDD>_<hhmmss>_<LABEL>_<HOSTNAME>_<ROLE>_<MODE>.tar.gz

Example 1) hsds_log_20260120_204236_labelname1_storagenode1_MasterPrimary_Base.tar.gz

Example 2) hsds_log_20260120_204236_labelname1_storagenode1_Worker_Base.tar.gz
When splitting of dump log files is specified:

hsds_log_<YYYYMMDD>_<hhmmss>_<LABEL>_<HOSTNAME>_<ROLE>_<MODE>.tar.gz.<INDEX>

Example 1) hsds_log_20260120_204236_labelname1_storagenode1_MasterPrimary_Base.tar.gz.001

Example 2) hsds_log_20260120_204236_labelname1_storagenode1_Worker_Base.tar.gz.001

Note:

When storing dump log files, ensure they are saved in paths with sufficient length to avoid violating path length restrictions specific to the OS you are using.

When executing each REST API/CLI in the procedure, authenticate with the issued authentication ticket and execute it. At the time of authentication, use the username and password that were specified when the authentication ticket was issued. If an authentication ticket has not been issued, issue it according to Issuing an authentication ticket in the VSP One SDS Block and SDS Cloud System Administration. If you cannot issue an authentication ticket due to a failure of VSP One SDS Block, contact customer support.

In VSP One SDS Block, the system might automatically create dump log files in the event of a failure. Verify whether dump log files have been created according to the procedure below, and if they were created automatically, download the dump log files. Automatically-created dump log files include "AutoCollection" for <MODE>.

A dump log file created with the REST API/CLI includes a value ("Base", "Memory", "Monitor", or "All") set in the mode parameter for <MODE>. When the mode parameter was omitted, "Base" is set.

Additionally, <ROLE> contains information about the role of the storage node targeted for collection of dump log files ("MasterPrimary", "MasterSecondary", or "Worker").

A maximum of one generation of automatically-created dump log files and a maximum of one generation of dump log files created with the REST API/CLI are stored in each storage node. When a new dump log file is created beyond the maximum generation, the old dump log files are automatically deleted (event log KARS10668-I is not generated at this time).

Note:

If you have not yet imported a server certificate into the VSP One SDS Block, you may receive a warning message about the server certificate.

If you see a warning message, check the IP address or host name of the request destination, then see Action to be taken when a warning message about a server certificate appears in VSP One SDS Block System Administrator Guide and follow the warning ignore method for each execution environment listed in the bulleted list.

Verify the dump log file creation status.

Perform this step for all storage nodes.

REST API: GET /v1/objects/dump-statuses

CLI: dump_status_list

If dump log files whose mode is "AutoCollection" exist in the storage nodes, verify the status of the dump log files.

When the status is "Creating", wait until the status becomes "Created" or "Failed." It will take a maximum of 60 minutes until the status becomes "Created" or "Failed."

After the status of all the dump log files whose mode is "AutoCollection" becomes "Created" or "Failed", verify the size and fileName of only the dump log files whose status is "Created" and the number of files into which such dump log files can be split, and then proceed to step 2.

The size of the dump log files are displayed (unit: MiB) in "size."

If there is no storage node in which dump log files whose mode is "AutoCollection" exist, proceed to step 3.
Download the dump log files whose mode is "AutoCollection" and status is "Created" to the controller node.
Perform this step for all storage nodes in which dump log files whose mode is "AutoCollection" and status is "Created" exist.

Secure free space equal to or more than the size confirmed in step 1 before downloading.

REST API: GET /v1/objects/dump-files/<fileName>/download

CLI: dump_file_download
Note:
- If you specified an index parameter, you can download split files whose file size is 400 Mib. When you want to download the dump log files separately, perform download for the number of files into which dump log files can be split.
- If a dump log file could not be downloaded using a VSP One SDS Block Administrator, collect it using a REST API or CLI.
Request the creation of a dump log file from the controller node to the storage node to which the dump log file is created.
Perform this step for all storage nodes.

Specify "Base" for the mode parameter.

You can run the request for creating the dump log files in parallel.

REST API: POST /v1/objects/dump-file/actions/create-file/invoke

CLI: dump_file_create_file

Note:
When you specify <LABEL> of the file name in the label parameter, observe the following.

Number of characters: Maximum of 64

Characters that can be used: Numbers (0 to 9), uppercase alphabet (A to Z), lowercase alphabet (a to z), symbols (! # $ % & ' - . @ ^ _ ` { } ~)

CAUTION:
- When failure information is being generated in the storage node, an error message (KARS10665-E) is returned. When this message appears, wait for a while, and then retry executing the request for creating the dump log files. If the message is consecutively returned for 60 minutes or more, skip step 4 and subsequent steps for the storage node.
- If the information to be collected does not exist on a storage node, the KARS10666-E message is returned. If this message is returned, skip the operations in steps 4, 5, and 7 for the relevant storage node.
Check the creation status of the dump log file.

Perform this step for all storage nodes.

It takes a maximum of 60 minutes to create the dump log file.

REST API: GET /v1/objects/dump-statuses

CLI: dump_status_list

When the status of the dump log files for which you requested creation changes to "Created", verify the size, fileName, and the number of files into which the dump log files can be split, and then proceed to the next step.
Download the dump log files to the controller node. Secure free space equal to or more than the size confirmed in step 4 before downloading.
Perform this step for all storage nodes.

REST API: GET /v1/objects/dump-files/<fileName>/download

CLI: dump_file_download
Note:
- If you specified an index parameter, you can download split files whose file size is 400 Mib. When you want to download files separately, perform download for the number of files into which the dump log files can be split.
- If a dump log file could not be downloaded using a VSP One SDS Block Administrator, collect it using a REST API or CLI.
Verify that all the dump log files you downloaded in step 5 were created with names conforming to the naming convention.
Delete the dump log files from all the storage nodes from which you downloaded them in step 5.

REST API: DELETE /v1/objects/dump-files/<fileName>

CLI: dump_file_delete

Note:
Running the above REST API or CLI does not issue a job. Event log KARS10668-I is output once the dump log file deletion is complete. Therefore, verify the event log.
If the CLI was also being run, collect the CLI log in the controller node as described in Procedure for collecting CLI logs.

Note:
When sending dump log files to customer support, give priority to those files whose <MODE> in the file name is "AutoCollection."