-
Notifications
You must be signed in to change notification settings - Fork 96
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Fix the errors in the dashboard for the AIO installation #2117
Conversation
Signed-off-by: Miguel Verdaguer Velázquez <[email protected]>
Signed-off-by: Miguel Verdaguer Velázquez <[email protected]>
Signed-off-by: Miguel Verdaguer Velázquez <[email protected]>
Signed-off-by: Miguel Verdaguer Velázquez <[email protected]>
…ct help commentary for argument --development Signed-off-by: Miguel Verdaguer Velázquez <[email protected]>
Signed-off-by: Miguel Verdaguer Velázquez <[email protected]>
Signed-off-by: Miguel Verdaguer Velázquez <[email protected]>
Test failed for Amazon Linux 2, 2GB 1 CPU coreWhile testing an AIO installation on an Amazon Linux 2 machine with 2GB of ram and 1 cpu core, an error has been seen:
The installation failed, but the retries of dashboard were called correctly. By stopping the installation after the last retry, it has been proven the wait done in the retries is not enough at all, as, after waiting a couple of minutes, the dashboard is initialized. SolutionAfter increasing the time between retries and the number of retries, the problem has been solved. |
Test failed for CentOS 9 Stream, 2GB 1 CPU coreA first try failed because of an error in the Wazuh indexer installation:
A second installation also failed, this time, because of the API, although it seems the changes of this PR worked correctly and the problem didn't come from the token download.
A third installation has replicated the second error, this time with more debugging:
This error doesn't come from the changes of this PR, and it seems the installation fails because no user is passed on to Some more tests have been done in a different and newly created CentOS 9 Stream and the error has not been replicated, as seen in the following trace:
Another test is going to be done with only 1Gb memory to try and replicate the error, as the main difference between the two tests was in the first one the machine was created on a host much more on the limit, memory wise. The test with 1 GB has also been successful, so probably the first error came from a virtualization issue or the previous host.
|
Description
This PR aims to fix the errors seen in the dashboard installation when an All-in-one installation is done in a system with less capabilities than the recommended ones. This errors have discovered some issues on the error treatment of the installation.
For now, function
passwords_getApiToken
now detects one more possible error and retries the Token downloading some number of times.Testing
Test AIO on machines with minimum requirements(4GB and 2 cores):
🟢 Ubuntu 22
🟢 CentOS 8
🟢 Amazon Linux 2
🟢 RHEL 9
Test AIO on machines with limited resources (2GB and 1 cores):
🟡 CentOS 9 Stream
Details of the error in commentary https://github.com//pull/2117#issuecomment-1462220195🟢 Amazon Linux 2
Solved a problem as explained in #2117 (comment)
🟢 CentOS 7
🟢 RHEL 7
🟡 Ubuntu18LTS
A first installation has failed with an error starting the wazuh-manager, very probably on account of a memory shortage
After some more installations have been done, it can be said when the manager starts, there is no problem with the API or dashboard, and so the changes in this PR work correctly.
A final test using a machine with only 1GB also finishes successfully: