Ashoat

Looks like the folks at cPanel have resolved the issue.

Closing due to inactivity...

It's looking pretty indefinite at this point. Currently, cPanel/WHM depends on OS vendors for PostgreSQL versions, and RHEL/CentOS currently has it at 8.1. No idea when they'll push 9.0... RedHat is usually pretty conservative with software versions. There's a cPanel feature request here for the company to maintain its own version of PostgreSQL.

I've submitted a support request to cPanel in regards to this issue, as I am unable to resolve it myself.

Ugh... this is annoying, considering that I already switched my scripts to use Rails 3.0. Let me see if there is any way out... Looks like one option is to get rid of Mongrel and use mod_rails instead. However, that will mean having to manually restart by creating a blank file, and having to do some configuration with .htaccess files. Alternately, could somebody having this issue try to add the following to their Rails gemfile? gem "mongrel" gem "cgi_multipart_eof_fix" gem "fastthread" gem "mongrel_experimental" Since we're planning to switch Rails accounts to Charlie eventually, I'm thinking that it might make more sense to avoid downgrading our version of Rails.

After a good bit of work I've finished the script to fix corrupt domain entries. FTN: you should be good now. Please wait until "Server Load" under "Service Status" in your cPanel is below 4, and then try adding the domain again once. You will have to wait 24 hours after adding your domain for it to become active. Everyone else: please repost your issue in the format described in the first post of this thread. Admins: with this new script, it's really easy for you guys to deal with these issues, so I'm deescalating this thread. To fix a corrupt entry, simply run: sudo /root/python/cli/fix_corrupt_domain.py username domain For example, for FTN you would do: sudo /root/python/cli/fix_corrupt_domain.py ftn friendlytechninja.com The above script will remove all entries for the given domain. Additionally, if it finds other corrupt entries for that user it will pick them up and delete them.

No problem. Sorry about the lack of notice. Closing thread now...

Actually, the total number of accounts on Stevie is currently closer to 25,000. But perfectly accurate on all the other points.

Yeah, that version is pretty outdated. Okay, updated.

I'm pretty sure that it's already at version 3. root@stevie [~]# rails -v Rails 3.0.3

Glad you get it working!

You need to /delete/ that file. The script Geoff wrote for you will delete the file for you.

Wizard: it's not something you can handle. When / is filled up it's usually due to /root being filled up, which is filled with my stuff. Closing thread...

Okay, should be working now.

If it says it's already configured but it's not working, you'll have to wait a bit for me to set up the scripts that broke two or three months ago. That's next on my to-do list, though, so it should be within the next few days.

I just made some important changes to a script that checks for corruption in our database. It now takes 500 MiB of RAM instead of 2 GiB and runs much quicker. Amazing how much difference the XML library you're using can make!

Well then it's already registered to you...

I'm looking at the domain record, and it seems to be owned by "ftn". Is that you?

Okay, signups should be open again now.

I guess I could confirm ownership of the domain by emailing the listed contacts. But does it really matter when the domain isn't even being used here?

Okay, everything worked! All accounts should now have unmetered bandwidth.

Are we talking about friendlytechninja.com? That domain's nameservers aren't pointed here.

Yup! I actually spent most of yesterday figuring out how to manually modify cPanel account files, and I wrote some scripts to fix everything up. At this point, I'm just waiting on a script to check and confirm that everything went as planned.

Not right now, but we should have signups up again within 24 hours.

Last two "News" items should explain it. The downtime over the last three days was a result of trying to change account settings, and the general instability over the last few months is the result of a software bug (now fixed).