Manager lost its styling and functionality

Hello folks,

Please see the attached screenshot of my manager. This is in Revo 2.2.7. 2.2.8 is doing this as well. This site is hosted on fatcow.com.

My manager was running fine on 6/5. On 6/6, the attached screenshots started happening.

When I log in fresh from the login screen, all looks ok. When I click my Home page to edit the content, the manager goes blank and the links show up after about 10 seconds as in the screenshot. If I do a refresh on my browser, the manager tree comes back ok, but everything to the right of the tree is blank. Same things happen if I click on a snippet or a chunk to edit.

I have looked at the following link and tried the indicated troubleshooting tips:

http://bobsguides.com/modx-troubleshooting.html

When I look at my site files via Filezilla, I do not see a folder titled logs under core/cache.

Could someone please point me to where I can find the log files?

Excerpt from Bob's Guides:

"If you can't log in to the Manager or if the Manager looks wrong (e.g., no menus, missing or empty Resource tree, blank panels, etc.), try changing the compress_js and compress_css System Settings. If you can access the Manager go to System | System Settings and type compress in the search box at the upper right. If you can't access the Manager, use PhpMyAdmin and make the changes in the modx_system_settings table. Find the two settings and change the value from 1 to 0, then click on the "Go" button."

I changed this in the database via phpMyAdmin, cleared my cache by going to core/cache and deleting all files. I cleared my browser cache. I sign in and click on my home page in the document tree, things now appear to be working.

Here are my component versions:
Iceweasel 21.0
Chromium 26.0.1410.43 Debian 7.0 (189671)
Hosted on fatcow.com
PHP 5.3.13
MySQL 5.0.91-log
Apache 2 (having trouble finding the exact version, will add when I find it)
Fresh installs of MODx Revo 2.2.7 and 2.2.8

When this initially happened, I wiped 2.2.7 off the server and reloaded my backups of 2.2.7 from several days before that was functioning fine when I backed it up. This manager issue happened on my restored backup files and database as well. I wiped everything and started with a clean install of 2.2.8, manager issue was still present. I went back to my backup copy of 2.2.7 to continue troubleshooting.

I also cleared out my PHP sessions just in case something was weird in there.

I thought it was very odd that things were working fine one day, then went south the next day. I have worked with MODx Revo on GoDaddy, Bluehost, Hostmonster, Host Gator, and Siteground with no issues like this.

Has anyone experienced anything like this where it works one day, breaks the next day, and fix it by switching a setting in the database or tweaking the install?

Thank you all for your comments and direction.

Hosting companies quite often make changes to their server configurations without any warning. I've had sites fail overnight, and sometimes have had to get quite rude with the hosting tech support to get them to admit that they did change something.

Unfortunately both of my clients on FatCow are using Evo, and they're both fine.

Thank you for your reply. I have been very fortunate as to not have had any issues up until now with changes the hosts make behind the scenes. I will keep this in mind when something goes wrong in the future.

Strange but true - I have a site where the manager does exactly this in every browser except Safari. No guarantees, but you might just give it a try ...

This sounds familiar. This could be a case where the hosting account is running out of memory and is unable to produce the CSS/JS files to run the manager. This is why it's so critical to turn off CSS/JS on servers with little memory. This could also be a permissions issue, however you're seeing 500 errors rather than 404s on the files.

When you mentioned memory, that lead me to check the memory_limit in the php.ini. It is currently set to 32M. I checked another site on a host I have used for several other clients and that server is set to 128M.

Fatcow was a bit different in that I had to manually set where the PHP sessions are saved, the timezone was left blank, and I had to choose between PHP 5.2 or 5.3. The other hosts I have used have had everything set automatically, so I did not think to check these settings upon my initial loading of MODx.

I am going to try and experiment being that this is only affecting the manager. I am going to set the memory_limit to 128M, turn the compression back on, and see what happens the next few days. I am the only one getting in the manager right now, so it is only affecting me on the backend. The front end was never affected on this site thankfully.

Thank you for your reply. I will post my findings in a few days.

CSS/JS Compression isn't really all that valuable in my opinion. I'd leave it off and see if you can isolate everything else that could be causing issues.

Just logged in to my manager today after turning compression back on and it is hosed again.

I turned it off, cleared my site and browser cache and all is well.

I turned compression back on to cause the error. I logged in and all looked good. I then clicked on the Home resource and bam...manager is hosed again.

Here is what was generated in my error log:

[2013-06-11 12:03:19] (ERROR @ /connectors/system/settings.php) Error caching lexicon topic lexicon/en/tinymce/default

I cleared my site and browser cache again, turned compression off, logged back into my manager, all is well.

I clicked on my elements tab to prepare to click on a snippet. I turned compression back on. I then clicked to edit a shippet, hosed again. Here is the error generated:

[2013-06-11 12:12:26] (ERROR @ /connectors/layout/modx.config.js.php) Error caching lexicon topic lexicon/en/core/resource

Again, turned compression off, cleared browser and site cache, all is well again.

You have a great point Jay. I would like to fix whatever the root cause is. Turning compression off definitely fixes my issue, however, when I installed MODx initially, everything was fine for about 10 - 12 days, then the manager issues started to appear. If this seems to be a resource issue, I can accept that and move on and make sure I turn the compression off in my future installs.

If by the info provided from the error log there seems to be an underlying issue, please let me know and I would like to continue to work toward a solution. I definitely don't want to make a mountain out of a mole hill, so please let me know if I should or should not pursue this any further.

Thank you again for your time and information.

Dan

When you're using the manager are you checking the network console in the browser dev tools? Are you seeing 401s or 404s on requests. Don't assume that because JS/CSS compression is on by default that it's the best option for your site. I advocate for that being off in it's current state on install until it can be optimized further. There is little reason to have it on in my opinion. I have been seeing more and more hosts doing silly things with memory allocation to individual threads and/or thread numbers running which can also cause these issues. It might be worth looking at the PHP/Apache logs if you can access them. Depending on your hosting panel you may also be able to see if you're running out of memory or is being throttled. I am not the best person to debug this issue but I am sure someone smart could help get to the heart of it.

Sorry for the late response. I will get in on Monday to recreate the problem and pull fresh logs. I will post the log info and any screenshots I take as well. Thank you again for the direction on this.