Describe the bug
A clear and concise description of what the bug is.
Clients are unable to connect over HTTPS, but once you change configuration to http / port 80 problem resolves. This began happening 5/30/2020.
Today (Saturday, May 30th) we found that none of our Libki clients were able to connect to Libki over https. Prior to this, NO changes to settings or updates were given to:
a. Libki
b. Apache2
c. Ubuntu
d. MySQL
e. Configuration files.
*We have a cardinal rule about not making system changes on Fridays unless you want to ruin your weekend! :)
The WebGUI was running fine, responsive, and all of the users, history, data were intact.
Libki's log files showed no client attempts to connect.
Our firewalls and routers showed no blocking events.
Diagnosis steps:
Our environment is fully virtualized (not just Docker) - so we tried
a. Rolling back to a known good configuration from a week ago. (Did not work)
b. Restoring from a full backup backup on May 13th (fresh build day). (Did not work)
c. Restoring our prior Libki (v.1901). (Did not work.)
d. Changing clients to use scheme http and port 80 instead of https/443. (Worked).
--Note that in this setting apache is still reverse-proxying Libki, just not over https.
Possible causes (unproven speculation)
A mandatory uncheduled Windows 10 update broke openssh on the clients?
To Reproduce
Steps to reproduce the behavior:
- Configure clients to use scheme https, port 443
- Restart clients.
- Clients vanish from libki and do not re-register.
Expected behavior
A clear and concise description of what you expected to happen.
Clients should be registering against Libki server in a timely manner over either http or https.
Screenshots
If applicable, add screenshots to help explain your problem.
Deployment architecture:
- OS: [e.g. iOS] Ubuntu 20.04 server
- Deployment style ( LAMP, Docker, etc ) LAMP
- Versions of software used Libki server 20.05 as of May 13th build status.
Desktop (please complete the following information):
- OS: [e.g. iOS] Windows 10 Professional 64
- Browser [e.g. chrome, safari] N/A - but clients can connect to libki GUI using any browser.
- Version [e.g. 22]
Additional context
Add any other context about the problem here.
Our most recent Windows updates (clients) were run Tuesday, May 26. We control when we deploy updates, but sometimes Microsoft knows best and forces an unannounced patch. Clients were connecting over https/443 through late Friday (may 29) with no issues.
Describe the bug
A clear and concise description of what the bug is.
Clients are unable to connect over HTTPS, but once you change configuration to http / port 80 problem resolves. This began happening 5/30/2020.
Today (Saturday, May 30th) we found that none of our Libki clients were able to connect to Libki over https. Prior to this, NO changes to settings or updates were given to:
a. Libki
b. Apache2
c. Ubuntu
d. MySQL
e. Configuration files.
*We have a cardinal rule about not making system changes on Fridays unless you want to ruin your weekend! :)
The WebGUI was running fine, responsive, and all of the users, history, data were intact.
Libki's log files showed no client attempts to connect.
Our firewalls and routers showed no blocking events.
Diagnosis steps:
Our environment is fully virtualized (not just Docker) - so we tried
a. Rolling back to a known good configuration from a week ago. (Did not work)
b. Restoring from a full backup backup on May 13th (fresh build day). (Did not work)
c. Restoring our prior Libki (v.1901). (Did not work.)
d. Changing clients to use scheme http and port 80 instead of https/443. (Worked).
--Note that in this setting apache is still reverse-proxying Libki, just not over https.
Possible causes (unproven speculation)
A mandatory uncheduled Windows 10 update broke openssh on the clients?
To Reproduce
Steps to reproduce the behavior:
Expected behavior
A clear and concise description of what you expected to happen.
Clients should be registering against Libki server in a timely manner over either http or https.
Screenshots
If applicable, add screenshots to help explain your problem.
Deployment architecture:
Desktop (please complete the following information):
Additional context
Add any other context about the problem here.
Our most recent Windows updates (clients) were run Tuesday, May 26. We control when we deploy updates, but sometimes Microsoft knows best and forces an unannounced patch. Clients were connecting over https/443 through late Friday (may 29) with no issues.