Poor manager performance

We are running MODX Revolution 2.1.3 on a web host. Unfortunately right now I don't have the hardware specs for that shared hosting. The software specs are:

Shared Unix Hosting
PHP: 5.2.9
mySQL: 5.0.77 / Protocol Version 10
Apache: ?
ModX: 2.13-pl

We have really terrible performance from the manager UI. It can take anything from 10s to 30s (and very ocassionally much longer than that) to save pages, update the tree etc. I get roughly 3 times faster performance from the manager on my local test install (which is still rather slow but usable)

So my question is - what kind of performance do people typically see from the manager UI? Is the speed of this a current achilles heel? Or do we have a problem?

On a hunch that it might be DB related, I ran some test queries and benchmarking - the shared host seemed to be many times faster than my local system for mysql queries so I guess it isn't that!

Yours is running exceptionally slow - I've just timed a manager page refresh (editing resource page) on a remote, shared server with 2.1.3-pl and the page was fully loaded in just under 3 seconds. Local version is usually less than 2s.

If your resource tree is expanded and you have a lot of items it can take a while to load up, however something is definitely wrong in your case.

Maybe try running it through Firebug / yslow and see what is taking so long to load?

Also, which browser are you using? The manager is JS heavy so maybe try it out in Chrome and see if it's quicker there?

Thanks for the reply. We have only a few pages in the tree and I use chrome. I don't suspect the JavaScript execution since it runs much better locally. Virtually all the time is taken server side. I'll try using the chrome developer tools to see exactly where the manager loses time.

Who is the hosting company and is the website also slow or just the manager?

Do you know if the database server is on the same server as apache? If you are using something other than localhost or 127.0.0.1 then it's on another server.

You generally get better performance when the website and database are on the same server, granted there are some usage cases for a dedicated database server, but that's another story.

Pretty sure the db and apache are in the same server. The host is genotec - generally we've had no issues with other php sites we've hosted with them. The website itself was also performing badly but that turned out to mostly be lack of caching for Wayfinder, once that was sorted it performs ok.

To update this thread, I spent quite some time with Chrome Developer tools looking at the page loading in the Revolution Manager. When I run locally the initial request for an edit page for a resource gets a response in about 1.5s, then entire page including tree is complete within around 6-8s with all the contained php requests returning in around 1-1.5s each. For either of our remotely hosted instances, that initial request takes 5-6s to return, and all the contained php calls respond in 2-2.5s. Does the page loading for a particular resource depend on the amount of content? There is more content in both the remote instances, and I can see how that would slow down the building of the tree but not the response to the initial page request?

All considered, you need a new host. More content = slightly more load time but not anywhere near that margin. If your error log looks okay, you should start shopping for a new host or ask your current host what the deal is.