[SOLVED] Waptexit problem: 408 client error: for url:

[wapt.user6325]

Hello everyone,

The short intro:
Former administrator of a Wapt Enterprise in my previous company, having recently changed companies, I decided to set up a Wapt Server (Discovery for the moment, and we'll see later for the rest).

The problem :
The server is up and running, it seems to be working, waptdeploy was sent via GPO, BUT I have exactly the same problem as in this topic, which is marked as resolved (following the WAPT update)
Topic Link

Basically, when the client fails to contact the server upon shutdown (in my case, I'm not yet in production, so I'm simulating this with the WAPT server shutdown), I get a waptexit error:

408 Client Error: for url:
http://127.0.0.1:8088/is_entreprise
Press "OK" to ignore and risk data corruption.
Press "Cancel" to close the program

Following the thread mentioned, I thought this problem was solved? Am I an isolated case?
A licensing issue?

Thank you for your help

PS:
I don't think this is related, but I noticed that on my WAPT server's webpage, to download the client, it says "WAPT Server: ENTERPRISE"
I don't know if it should say something else, like "Discovery" since there is no license installed?


Test environment:
Wapt Discovery
WAPT Server version: 2.2.3
WAPT Agent version: 2.2.3.12463
WAPT Deploy version: 2.2.3.12463
Server OS: Debian 11 - up to date
Waptdeploy client OS: Windows 10

[alejeune]

Hello,

we have identified the cause, and an internal support ticket has been opened.
For a quick fix, you can add square brackets [] to the licences.json file on the WAPT server and to /var/www/licences.json on your Debian 11 server.

Regards,
Amélie

[wapt.user6325]

Hello,
Yes, that seems to have solved the problem! Thank you
(modification made, server rebooted, and no agent regenerated).

For my own understanding:
I don't quite understand how a modification to a server configuration file will affect the behavior of waptexit (which seems to query the "localhost" address, but it's incorrect) on the client machine?


Have a good day,
Alexis

[dcardon]

Hi Alexis,

the agent needs to know if it's running the Enterprise or Discovery edition to determine which features to activate. To do this, it retrieves a derivative of the license and stores it in a local cache. In the case of Discovery, there's no license on the server, so it should retrieve an empty JSON object (hence the brackets [ ]), but there was a bug where it was retrieving a null value instead. If the value is null locally, the agent assumes it doesn't yet know if it's running Discovery or Enterprise, so it connects to the server to check. This is the problem when the machine is disconnected and can't reach its server. The problem only appears when there's no ICMP unreachable response to close the connection (basically, there's a DROP). The agent then waits for a response that never arrives, and since the timeout between the agent and the server is longer than the time between waptexit and the agent, it returns a 408 code to waptexit.

The problem is convoluted, which is why it took us a while to figure it out. Basically, the server should never have returned a null value; it should have returned an empty list (the file you created).

Since the Enterprise version is what covers salaries and WAPT development, we naturally spend more time developing and testing it. This bug had escaped our notice.

Sincerely,

Denis

[wapt.user6325]

Hello Denis,

and thank you for your reply, which certainly satisfies my curiosity!

Have a good day and best of luck!

Alexis