1

I have a client site that was updated to ee2 (2.5.0) about 15 months ago. Its at http://www.smart-publications.com. The host both after and before that upgrade is yourhost.com. Everything worked fine. I updated them to 2.6.1, and updated all the add-ons that needed an update, about 10 days ago. Deployed the updates and within 15 minutes the server came down due to runaway memory usage. After working with the host and trying a few more times, no change.

We run a Rackspace cloud sites server for some client hosting, so I setup a site there and deployed the files. Using the testing URL that Rackspace provides, the site seemed fine. I opened the site in 8 tabs with auto reload on and the memory didn't go nuts. So I pointed the A record to this server for the time being and that is where the site is running now.

I figured that Yourhost may not have servers optimized for EE so I recommended to the client setting up a VPS on Liquidweb, which I used for years with EE1 and 2 sites before Rackspace. We did that, and I deployed the site there and, using a spare domain name I own, set the site up for that. Opening several tabs with the auto reload running sent the memory on the server out of control.

I thought maybe it was a 2.6.1 issue, or one of the add ons that I had updated, so I blew away the site and redeployed it using the original 2.5.0 version before I had done any of the upgrades. Same problem.

Can anyone help be figure out what the hell is going on? I've never seen anything like this.

Improve this question
3
  • Note that trying all kinds of different templates on the site caused the same issue- it doesn't appear to be a single template.
    – Chad Crowell
    Commented Jun 10, 2013 at 5:28
  • I don't have an answer for you, Chad, but I recently saw a spike in memory usage after migrating an old 1.6 site to 2.6.1. One of the baffling issues was why the same traffic wasn't locking the server for the old site – expressionengine.stackexchange.com/questions/11067/… – I'd love to hear if you find any issue with 2.6.1 itself. I have a 2.5.5 MSM site with no such issues.
    – Mark J. Reeves
    Commented Jun 10, 2013 at 15:47
  • ...I have a 2.5.5 MSM site that gets more traffic, is on the same server config and has no such issues. The variable here is a lot of spam traffic, but that didn't take down the 1.6 version of the site.
    – Mark J. Reeves
    Commented Jun 10, 2013 at 15:55

1 Answer 1

Reset to default
1

After much rolling back, redeploying, branching, merging, and banging my head, looks like it was something to do with the FocusLab portable configuration setup. I set this up locally on the site after I FTP'd it down from the host but before I made my first GIT commit, so a full rollback and deploy didn't undo this part of the system.

Note, we use FocusLab portable configuration on many sites, and it works perfectly. I don't know how yet, but I screwed up its implementation when retrofitting it to the site as part the upgrade project. The site is currently running well with a standard EE config, and I'll find where my error was over the next few days and reimplement it correctly.

Improve this answer

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.