Ghost process exited with code: 0

[ghost]

This issue is a

• Bug Report
• Feature Request

Summary

Debug Information:
    Node Version: v6.11.2
    Ghost-CLI Version: 1.1.2
    Environment: production
    Command: 'ghost start'
Message: Ghost process exited with code: 0
Stack: Error: Ghost process exited with code: 0
    at ChildProcess.cp.on (/usr/local/lib/node_modules/ghost-cli/lib/utils/local-process.js:47:24)
    at emitTwo (events.js:106:13)
    at ChildProcess.emit (events.js:191:7)
    at Process.ChildProcess._handle.onexit (internal/child_process.js:215:12)

Ghost won't start - just emits this unhelpful error message!

Steps to Reproduce (for a bug report)

Not sure. Usually happens after an upgrade and reboot on the server. NPM packages are rebuilt at the same time but this doesn't always fix it.

Technical details (will be automatically output by Ghost-CLI if an error occurs):

• OS: Ubuntu 16.04
• Node Version: 6.11.2
• Ghost-CLI Version: 1.1.2
• Environment:
• Command: sudo -u ghost -H ghost start

Kate Note: Error message should be improved first. Try to improve other error messages as well.

[togume]

+1 on this. Happened in my local macOS build (I use for local writing). Did ghost upgrade and got this issue. My nvm works well, and I can create new local ghost instances without issues.

[acburdine]

@newey01c @togume would both of you be able to try running ghost run (@togume add the --development flag because you're running it locally) and see if anything is output?

If this happens again too you might try running ghost log just to see if there are any error messages.

I'll try to add some additional logic in here to output exactly why ghost stopped - although if there is no logging output there's not much that I can do unfortunately.

[ghost]

There is no helpful logging output unfortunately! If there was, I'd have dug in a little bit deeper trying to diagnose the problem. I think it might be to do with knex-migrator using a local copy of the sqlite3 npm module which then doesn't get recompiled on a node upgrade (and then ghost-cli doesn't properly report the error) - but I'm not fully sure yet.

[togume]

@newey01c - sounds like what happened to me...

@acburdine - I'll try that next time to see what happens. What I ended up doing was - getting the sqlite3 to work with the local db that threw the error, doing a db dump, upgrading everything, importing the dump.

[whitfin]

@newey01c that appears to be accurate. I had the same issue; cd current && npm rebuild fixed it.

[Khangeldy]

In my case, this actually open my site sudo nodemon --watch content/themes/[theme name]/index.js -e hbs,js,css. But ghost not updates on changes. Another problem appear when i assign owner of content/ folder to ghost user. My Atom could not save file after changes. In this case chmod -R 777 resolve it. Then gulp throw error with permission issues. So i used sudo with it. Maybe create more elegant way to start nodemon without any trouble? IF exist reply to me. PLEASE.

[jeffreywyman]

I found that my site URL in /opt/bitnami/apps/ghost/htdocs/config.production was missing "http://" and when I changed that I resolved the error by re-running ghost update --force

[vikaspotluri123]

@Khangeldy when it comes to theme development, your best bet is to install ghost locally (ghost install local) and run it locally (ghost start -D). Any changes you make to your theme will show up because caching is disabled. When you install locally, you don't need to mess with permissions - you will be running the process as your own user so permissions issues shouldn't be a thing.

I'm not sure I understand how this relates to the original issue, but feel free to correct me if I'm missing anything. If your issue doesn't specifically relate to the original issue, please check out the #help channel in the Ghost Slack chat if you have any questions! If you're looking for guidance on ghost install local, check out the local install guide.

Also, as a side note, please make sure you update the CLI to the latest version! There were some dependency changes and production improvements that will make your life easier 😜

[Khangeldy]

@vikaspotluri123 I am getting same error like this issue. When i start ghost start -D i got Ghost process exited with code: 0 . I am already installed ghost locally. And it's tutorial saying that to create ghost user and gain folder owner to him. Under developing theme section .

[kirrg001]

@acburdine

This looks like a mixture of reports.

I think it might be to do with knex-migrator using a local copy of the sqlite3 npm module which then doesn't get recompiled on a node upgrade (and then ghost-cli doesn't properly report the error) - but I'm not fully sure yet.

This should be fixed!?

I found that my site URL in /opt/bitnami/apps/ghost/htdocs/config.production was missing "http://" and when I changed that I resolved the error by re-running ghost update --force

Not sure how it's possible to enter a URL without protocol, there is a url validation. If it was manually edited cc @jeffreywyman, the CLI could detect a incorrect url format.

---

@Khangeldy Hard to guess why you are running into the problem. Could you double check the comments and see if any of the suggestions fixes your problem? Share as much as you can and upgrade to the latest CLI version.

[kirrg001]

• Error message should be improved first

[restuarifp]

I have similar issue with the same error message. In my case, when I rename folder name other than "ghost". When I rename it back to "ghost", everything goes back to normal.

• OS: Mac OS 10.14.1
• Node Version: 8.12.0
• Ghost-CLI Version: 1.9.8
• Environment: Development
• Command: ghost start

[pfist]

I encountered this problem today when I attempted to upgrade Ghost from 2.6.x to 2.8.0 before upgrading Ghost CLI. After trying a few different things, running ghost update --force appears to have solved it.

Worth noting: When I ran ghost update --force the last step said "Restarting Ghost" even though the instance didn't appear to be running. Might be related.

Debug Info

• OS: Microsoft Windows, v10.0.17763
• Node Version: v10.14.2
• Ghost-CLI Version: 1.9.8
• Environment: development
• Command: ghost start

[shriker]

I ran into this today and it was due to my current/ symlink being broken. Running ghost update --force fixed the broken link.

[chrishawn]

OK so i was able to fix this on macos by doing
brew uninstall --force binutils
more on how/why here: https://stackoverflow.com/questions/54014971/macos-dyld-symbol-not-found-usdt-create-provider

[chansecampbell]

This is still an issue when the folder name is changed after initialising

[brownpl]

Not sure if this helps, but if you are running locally, after I renamed the install folder and I needed to also review my config.development.json file as it was still pointed to the old path for the sqlite file.

[ericgopak]

Happened to me too, after I've copied a production config file to my local Ghost instance (before that I only had development.config.json).
First I was surprised by tons of complaints regarding file/folder permissions which I didn't have before, and fixing those permissions didn't resolve the issue.

What helped me was to simply run ghost start --development instead of my previously run ghost start). Now it's working again.

P.S. My setup is on Windows 8.1 with Cygwin, using node v10.13.0 with nvm

P.P.S. Most likely unrelated, but I've also run npm install from the /current folder before it started working again. Just in case I broke something somewhere