Just upgraded to [ 2022.2 (Build 8552) works fine for now on Google Chrom, but deploy just gives a 404 when done on MS Edge. Same user testet on both Browers:
GoogleChrome OK
MS Edge: 404
Created a new relase for a given channel and then pressed for deploy: 404 on Edge.
We use Edge for external users via a partner portal, internally we may use Google Chrome.
Thank you for contacting Octopus Support and welcome back to the forums!
I just set up a test instance of 2022.2.8552 on Windows Server 2022 with Edge Version 108.0.1462.54 (Official build) (64-bit) - so same edge version as you are running.
The Octopus UI has loaded fine with no 404 errors, I then created an environment, created a test project to run a script on the Octopus server, created the release and deployed it with no 404 errors.
I do not have a windows 10 machine to test this on but I imagine it will be the same outcome since I am using the same Edge version as you on my Server 2022 box.
Unfortunately this looks to be environmental, you mentioned:
We use Edge for external users via a partner portal, internally we may use Google Chrome.
Can you get me some more details on this issue please so I can see if I can help further:
Was this working before the upgrade.
Is the 404 happening when you load Octopus or is it only happening on deployments.
What does the 404 error actually say.
Is it happening on all projects when pressing the deploy button or only one project.
You mentioned its happening for the same user, is it happening for all users or just one, if its just one can you please get them to clear their cookies and cache on Edge and see if that makes a difference.
If its just one project for all users can you please get us a HAR file for a user when clicking on the deploy button (a HAR file in Edge not working and a HAR file in Chrome working would be great to have for comparrison if you are able to get us both). You can send it to us using our secure uploads site here.
Let me know once that has been uploaded as we are not notified when a user uploads to our secure site.
Thanks for getting back to me, I just tested deploying to a channel too and it does work in Edge without a 404 error so unfortunately I cannot reproduce this from my end.
Let me know how your testing goes and I can see if I can help,
Hey @user12 sorry for the double post but I just did a bit of googling of the error message:
unhandled error when communicating with the octopus server. The server returned a status of 404
…and found this on our forums. This user is having the same issue but with the releases page. Turns out they use IIS as a reverse proxy and it seems like IIS doesn’t deal with encoded urls correctly.
I am not sure if you use IIS as a reverse proxy or this find would be similar for your instance but considering this works in Chrome (internal) and not in Edge (external) id warrant you have some sort of proxy or firewall in play here which may be the cause of this issue.
Just thought you might like to see that forum post as it may help with your testing and troubleshooting.
Thanks for letting me know the release is now not working on any browser, this looks like an issue with the project ODSLoader.
Can you confirm it is just the release for that project and not any other projects.
If you create a new release for that project and deploy it does that work?
Are you able to click into the release you are trying to deploy that is giving you that error and does it load the page for that release?
Are you able to get us a copy of your Octopus server logs please and a HAR file when you try and deploy that release, if, when you click on the releases page it throws the 404 are you able to also get us the HAR file for that too please.
You can use the secure link I created earlier to post those up once you have them.
Let us know once they are uploaded and I will take a look for you,
Good to note rolling back fixes the issue, this is very strange, my guess is we changed some code for 2022.2 (it is a big major version you are jumping to from 2021.1 to 2022.2 - this is supported though)
We have a comparison page here which details all the breaking changes between your version of 2021.1 to 2022.2 latest.
In there I can see these things relating to releases:
7755 - Fixes an issue with the “Create Release” page for projects that rely on an image from a DockerHub (<https://index.docker.io>) feed
7585 - Fix NullReferenceException being thrown when creating or viewing releases with a version mask
7402 - Bugfix: Loading DeploymentTemplate for release created prior to project getting converted to vcs may throw 404
Those are fixes though for bugs so they have been fixed in 2022.2 latest which you went to.
Did you manage to get a HAR file? We cannot proceed any further now without one really as I need to see what happens when you see the 404 error.
Instead of a Server Log file can we get a system Diagnostic report instead please as that will show us a lot more than the Server log file.
Let us know when those are uploaded and we can take a look, you will need to upgrade again to 2022.2.8552 to get us those files unfortunately but its the only way we can dig into this for you.
Sorry to hear you have struggled with this upgrade process, the HAR file and some logs are needed unfortunately so we can investigate this for you but if you are happy to stay on 2021.1 for now (this is no longer supported as an LTS version just to let you know so we do not patch it anymore with bug fixes etc.) then that’s ok.
In future we have some documentation here on creating a test instance for upgrades / to test DB edits etc.
The HAR file documentation link can be found in one of the links I posted above but I have linked it in this reply so it is easy to get to.
You mentioned:
I guess in a cluster setup there is a common Database and not providing the needed test environment.
Do you mean you are running an Octopus HA environment? If so then yes each Octopus node will need to be upgraded one by one whilst in maintenance mode, once all nodes have been upgraded you can then turn off maintenance mode and you should be good to go.
We have a full guide here on upgrading an Octopus HA setup. In that guide there is a process for creating a test instance for testing the upgrade before you perform it live. This can also be used for a single node setup if you are not running Octopus in a HA environment.
Hopefully those links help you, do reach out in future if you are struggling with the upgrade as I am keen to get you upgraded safety without issues.