During the installation, deployment, and maintenance of the GBase database (GBase数据库), some error messages may be overly verbose or unclear, making it difficult to troubleshoot. Additionally, improper operations can cause cluster abnormalities.
This article summarizes some GBase 8c issue cases and solutions for sharing. If you encounter additional problems, please feel free to reach out to the relevant technical support personnel.
Issue 1: Dependency Causes Deployment Failure
Error Message:
Environment Check:
Run the command cat /etc/os-release to check the operating system of the machine, which shows it is UOS 20.
Error Analysis:
The error occurs because the installation process cannot connect to libldap-2.4.so.2, which causes the installation to fail.
Solution:
There are necessary dependency library files for the GBase 8c database in the package/dependency/uniontech path. Due to compatibility with special operating systems, the required library files are provided in the installation package, but adjustments are needed manually. Follow these steps:
1) Create a temporary directory temp2, copy xxxx.om.tar.gz to the temporary directory, and extract the tar package.
2) Go into the dependency directory.
3) Copy all files in uniontech to dependency/lib64.
4) Return to the temp directory, delete the original .om compressed package, and execute:
tar -czvf xxxx.om.tar.gz *
5) Replace the original .om extracted package with the newly compressed one. Then delete GBasexxxx.centos7.8_x86_64.tar.gz, and repackage it with:
tar -czvf GBasexxxx.centos7.8_x86_64.tar.gz *
After completing these steps, use the new installation package, extract it, and complete the installation deployment.
Issue 2: Port Display Problem
Error Message:
Error Analysis:
The error message indicates that the port is occupied, but upon investigation, no processes are actually using the configured port.
Solution:
The hostnames of multiple machines are the same. Changing each machine to a unique hostname resolves the issue, allowing installation and deployment to complete successfully.
Issue 3: Component Status Abnormality
Error Symptom:
Some nodes fail to start normally and are in a "dropping" state, for example:
Preliminary Steps:
After restarting the node_1 machine in the DCS, certain components show abnormal statuses.
Problem Analysis:
An abnormal etcd state impacts the synchronization of information among components, indicating an issue with the etcd component.
Solution:
Upon investigation, it was found that the firewall re-enabled after the machine restarted, causing cluster state abnormalities. Disabling the firewall resolved the issue, restoring normal cluster operations.
Summary
This article covers issues related to dependencies, ports, and component abnormalities. For other potential issues that may arise during GBase database (GBase数据库) deployment and usage, feel free to share in the comments.
Top comments (0)