JamesThorpe

Yep, it had git-repo-sync. I renamed that and now the sync is proceeding (though it's now complaining about branches with more than one head!), but at least we're past the initial issue, thanks!

Hi, We're attempting to synchronise one of our repositories with git (in this case, an empty repo in gitlab). We've already successfully done another, and I've previously synced other plastic repos with git too, we're just having issues with this one. When starting the sync, we see this error: We did indeed originally synchronise a MergeBot repo from the github source, but it's not this repo (and not even this server - we migrated from one plastic server to another a while back). I've seen a similar issue before that essentially said to go to %appdata%/Local/plastic4/sync/git and remove the content there, but I've tried that and still see the same issue. My plastic client was also a bit out of date, but I've just updated to 9.0.16.4851 with no luck. Any thoughts on what to try next? Cheers, James

Glad you got it working! Since I posted this, we've started building a new dev server. Like our old one, it's hosted in AWS. However to "fix" this problem (more like work around it entirely), I've used AWS target groups and a load balancer, coupled with subdomains of the server domain. Now we have eg https://plasticadmin.server.ourdomain and https://jenkins.server.ourdomain set up in Route 53 pointing at the load balancer, and the load balancer listener rules and target groups take care of routing to the right http port on the server. It doesn't load balance anything since there's only 1 server behind it, but it makes this stuff trivial

Hi, We've finally updated to a newer version with the web admin, and would like to make use of it. We have things like Trac and Jenkins running on this server also. Each one of these, as with Plastic, has their own inbuilt web server, so instead of having a myriad of ports open we instead use IIS to front them all and rewrite the URLs. This works fine for both Trac and Jenkins, as they have options that let you specify exactly what the external URL of the system looks like. I've not found a similar option for Plastic however, so when we attempt to access the box via, eg http://our.server/plastic, it successfully responds with the 302 to take us to the login page, but because it's unaware that it was not the root of the website, that 302 sends us to http://our.server/account. Similarly if I attempt to directly open http://our.server/plastic/account, the page doesn't render properly as all the other resources requested after the initial page load are all requested from the root of the website, rather than relative to the page. I could probably fix this with further rules in IIS to modify the responses to take this into account, but I'm wondering if there's a simple option I'm overlooking that would let me just put in the root url into the config somewhere, in the same way other tools can. I've tried modifying the "override host name" option on the Network configuration in web admin, but it doesn't appear to apply to this. Cheers, James