In rare case, on EVO platform use login and logout immediately, MGD might generate a log: UI_LOGOUT_EVENT: User '(none)' logout

user@router> show log messages |match UI_LOGOUT |match none
Apr 24 15:26:10 router mgd[25681]: UI_LOGOUT_EVENT: User '(none)' logout

[vrf:none] user@router:~$ last
......
user pts/0        10.10.10.10     Wed Apr 24 15:26 - 15:26  (00:00)

In normal case, user login and logout messages is as below:

Apr 28 08:09:08  router mgd[30770]: UI_AUTH_EVENT: Authenticated user 'user' assigned to class 'super-user'
Apr 28 08:09:08  router mgd[30770]: UI_LOGIN_EVENT: User 'user' login, class 'super-user' [30770], ssh-connection '10.1.1.118 60192 10.2.2.144 22', client-mode 'cli'
Apr 28 08:09:08  router mgd[30770]: UI_LOGOUT_EVENT: User 'user' logout

The mgd UI_LOGOUT_EVENT log: User '(none)' logout log generates in case of a user login and logout immediately (in a few micro seconds), which might abruptly kill the CLI session.
When we open the CLI session, the CLI needs to set some basic parameters such as username, login time, and tty. These parameters are used for logging the login and logout events by the management daemon (MGD).
The problem arises because before the CLI sets these parameters and sends them to the MGD, the SSH session is terminated and the CLI is abruptly killed. This results in the username appearing as "none" in the logout event message.
Add the sleep for 1 or  2 sec as per device load that will make sure the processes complete their execution and don't get abruptly killed due to closing the SSH session.