I know I'll have to rebuild all my microservices images myself, but is there any way to recover the default images from openshift?
For example:
image-registry.openshift-image-registry.svc:5000/openshift/nginx
Under the openshift-image-registry project, there's 3 deployments:
cluster-image-registry-operator
image-registry
registry-nginx
Would restarting any of those do the trick?
I wanted to delete all images from a specific project, but because of an error in my logic, the grep filtering the project didn't work and so it started deleting all images in all projects.
By the time I realized what was going on it had already deleted maybe 70% of the whole registry.
Related
I have a folder containing image files in wwwroot/images. Web Publish has stopped deploying them.
They're added to the project. They're in source control. But the build has stopped deploying them. Not happened before.
How do I make this software do what it's supposed to do?
The project file had accumulated some junk. For example, it had an odd selection of files as 'Content' and it was removing the images and adding them back not as content.
I deleted all of it. Mwahahaha.
Our release bundles are kinda huge (~50MB) and the clients take approx. 1-2 mins to download the update from codepush.
We're mostly just updating the js bundles, is there a way to just delta update the js bundle without re-downloading all image assets?
I've read from here that the server performs files diff, but it doesn't seem to work on my case: Logging RemotePackage.packageSize from checkForUpdate on clients shows the exact size of JS bundle + image assets, despite the fact that I've only changed the JS code.
Is there something I can do about this?
The CodePush backend should only serve the full contents up to the point it finishes creating a diff zip. On the backend your package is downloaded, extracted and analyzed for files that are different. Any new/updated/removed files are reconciled, put into another zip and shipped back up to the backend so users get served that instead of the original one you release. If you are seeing the full download, it's either because the backend is still working to process the diffs or there are differences inside your bundle.
I would try to do a release, wait 10 minutes and check to see if it is still downloading the full bundle. If it does, it'd be best to contact support for a deeper dive. As far as I can tell bundle's are processing correctly with next to no errors and running a quick test myself produced the correct results for me.
I am having the most frustrating issue with VQmod. I moved my OpenCart store from a Godaddy VPS to Rackspace's Cloud Sites. The move went fine and everything works properly except the VQmod's on the admin panel. None of them load. No errors in the log files, no admin cache files in the cache dir, no php errors.
Things I have tried:
Cleared all cache Changed admin folder to 755
reinstalled VQmod tried both manually and using the installer with fresh index.php files
Removed all XML files and tried to load only one at a time
Cursed loudly at my computer repeatedly.
Please Help! OC version 1.5.6 VQmod 2.5.1
For those that want the solution to this, the issue was that the config.php files were both using relative paths instead of the full paths for OpenCart's various directories. They should always be full paths, or resolved with realpath() in the config.php files themselves
My case was a bit different. I checked the permissions, paths, all the regular stuff that comes to mind first. I even walked step by step through the manual installation guide.
The Opencart copy in question is shared across several environments using git. Long story short, the mods.cache and checked.cache were not added to .gitignore right away, and when I finally did that, I emptied both of them just to make sure Opencart will write new content based on my current environment. Turns out, since mods.cache was empty, Opencart believed there are no mods available.
Solution: delete both vqmod/mods.cache and vqmod/checked.cache.
Update: here are some similar issues:
https://github.com/vqmod/vqmod/issues/32
https://github.com/vqmod/vqmod/issues/3
The vqmod/vqmod/wiki/Troubleshooting guide, as of now, does not make it obvious the files should've been deleted, neither does the vqmod/vqmod/wiki/Installing-vQmod-on-OpenCart, and there doesn't seem to be any way to contirbute. vQmod fails silently, without producing any notifications, warnings, or simply detecting the issue and rebuilding the cache files. I've spent few hours trying to figure out what's wrong.
I am working with Rails 3.0.3 and Bluepill 0.0.51. I am trying to troubleshoot a situation where Bluepill is trying to start multiple instances of my ruby servers in some cases, but having trouble knowing where to start since the only files I have to look at are my ruby server .rb files and my .pill file in my rails app root folder. What are the configuration/other files that Bluepill creates on my system? Thanks so much in advance.
When you do run a successful pill recepie it can create different files based on your configuration. If you demonize the process it will create a .pid for each process that stores a value that represents the Process ID of that process so it can shut it down. It will also create a log file and sometimes files in sock. It uses /var/run/bluepill by default but if you run it in --no-privileged mode it will want you to specify the location, preferebly in your applications folder.
It doesn't create anything to configure but rather you configure everything in the pill file. It can be tricky getting everything to work but you have to keep trying. I realize this post is old so i hope your solved it :)
My application runs as a Windows service. During normal execution, some data files are generated in a "data" directory that I create with my WIX install. During an upgrade, I would like to move/migrate those files to the same data dir in the upgraded installation. However, I am running into issues. I am using the "MajorUpgrade" tag as follows:
I tried scheduling later in the process(afterInstallFinalize), but that runs into issues with the service having locks on files during the upgrade. I have combed Stack Overflow and google, and it seems that no one else is describing my scenario. Others have default *.ini files, that the user can potentially edit. However, in my case, it should always be safe to migrate the files, as there are no defaults. They are simply data files that need to be migrated.
I thought I could possibly do this via a custom action, but was not sure how to do it.
Any suggestions would be very much appreciated.
Edit: Apparently using "afterInstallFinalize" works fine for me, i.e. data files are migrated, as long as the user does not change the path during upgrade. If I change the path during upgrade, I get warnings about files being locked, and asking to stop the app using the files. However, it ends up not migrating the files in that case.
Ultimately, I solved my own issue by simply overriding the default dialog behavior, and changing the flow such that on upgrade, I skip over the "InstallDirDlg".