Has anyone had a problem with the Vision Clients page on the gateway web server? Mine just sits there “Loading…” but nothing ever shows up.
Yes, we have that issue as well. [with 7.9.1rc]
In addition - we get very high CPU usage on the webserver thread when activities like trying to view the audit log timeout.
We are using the internal Ignition database for select queries only. I’ve been wondering if we are creating some deadlock scenarios with the SQLite connection. You don’t by chance also have a connection established to the internal db, do you?
Nope. I am not touching the Ignition database.
Working now… IDK.
Ours also works on and off.
I noticed in the 7.9.1 official release notes some items related to the webpage. Maybe this was found and fixed?
We are seeing this on one of our gateways as well.
And the longer you let it sit there with the LOADING… the more memory the gateway consumes.
Has there been any insights on this?
I am again having this problem. I use a login script to check if a the current user already has a project open so we don’t have multiple clients running on the same machine. Sometimes a client can’t start a project because the gateway seem to think they have it running. When I go to look at the list of clients with the intent of stopping the client I can NEVER get the client list to appear. There seems to be a connection between these two conditions.
How are the clients tracked? Is it possible that when a client closes it does not correctly remove itself from the client list? Is it becoming corrupted to the point that the clients page to not work?
Is anything visible in the browser console (F12) when you navigate to the Vision Clients page and it doesn’t respond?
Not sure. I assume that the data is loaded dynamically into the page. It is now working and the client who could not start a project now can. I believe they are somehow related.
I also got this problem time to time. Is it something that can be fixed? I’m running Ignition 7.9.3.
This has become a problem for me as well.
The Vision Clients web page sits there loading and eventually crashes.
I have also found the same correlation to this occurring when it thinks a project is open that is not.
I have had to restart the gateway every time this has happened.
Was there ever a solution found?
(I am running v7.9.6)
There are a couple of possible causes. If you’re having crashes and you don’t have one of these situations, please contact support, because you’re experiencing something we don’t know about:
Read timeouts on SSO were able to cause zombie client sessions. A flaky enough connection could lead to thousands of these zombie sessions, which only went away on gateway restart. The client status page was trying to parse information about all these zombies, and ran out of memory to do the task. This is fixed in 7.9.7
If you have enough legitimate client sessions and not enough browser memory, the page loading would time out. We changed all the pages were this could possibly happen to use paged tables, which has solved that problem. This fix is in 7.9.9
I like the term zombie session and this makes sense for my problem. As I said above I have a script that prevents more than one instance of a project to run on a single machine. If the client does not shut down “gracefully” the client will not get removed from the “active clients and what they are running” table. Then I get a call saying someone can’t get start a project and the first place I go is the list of clients. Which is when I discover the problem. I will make sure we are at 7.9.7.
If you figure out how to do this right now, please let the rest of us in on the secret handshake… (-:
V7.9.9, Dang. Hoping it would be V8
I’ve arranged for a time machine and now this fix is in 7.9.8. Bwa ha ha ha!