Argh, the Umbraco forum ate my reply (as it sometimes does, annoyingly), so I'll reply again more briefly:
Disable your dashboards (see config/dashboards.config).
Do you have uCommentsy/uBlogsy/Archetype installed? I've had issues with them.
What plugins aside from UrlTracker do you have installed?
Have you tried tracing with dotTrace?
What version of Umbraco do you have?
Are you doing anything fancy (Azure, load balancing, network folders, etc.)?
You can configure SQL Server so it uses less memory.
Is your bottleneck CPU, Hard Drive, or Network?
Anything strange in your Umbraco logs?
This pegged CPU has been happening to me in Umbraco 7.1.8 an 7.2.2, so maybe it's happening to you too (if so, I encourage you to vote for it): http://issues.umbraco.org/issue/U4-6292
Disable your dashboards (see config/dashboards.config).
will try this and see if it makes the dev site faster
Do you have uCommentsy/uBlogsy/Archetype installed? I've had issues with them.
Yes. would it do good if i upgrade?
What plugins aside from UrlTracker do you have installed?
Active Directoyr Backoffice Users
Admin Session Manager
CmsImport
CMSImport RSS Data Adapters
Dewd
Diplo Link Checker
uBlogsy
uCommentsy
uComponents
Umbraco Contour (paid)
Url Tracker
Have you tried tracing with dotTrace?
no. dont i have to have a VS Proj for that to work? we just install what we get from umbraco version packages
What version of Umbraco do you have?
Currently have 6.2.4 for production. Dev and Staging both have 6.2.2. all 3 are slow.
Are you doing anything fancy (Azure, load balancing, network folders, etc.)?
No. we are considering load balancing to avoid the site crashing. the production/live site keeps crashing
You can configure SQL Server so it uses less memory.
thanks
Is your bottleneck CPU, Hard Drive, or Network?
Figuring it out.
Anything strange in your Umbraco logs?
No.
This pegged CPU has been happening to me in Umbraco 7.1.8 an 7.2.2, so maybe it's happening to you too (if so, I encourage you to vote for it):http://issues.umbraco.org/issue/U4-6292
Definitely disable some of your dashboards then. In particular, I have had major problems with the uCommentsy dashboard before. Whenever I logged in, that dashboard would show up, and it would take a minute or two for the backoffice to load. It had some seriously slow code that I think may have been optimized a bit (I submitted a pull request, but I don't remember the details). Upgrading might help, but I'd definitely start by disabling the uCommentsy comments approval dashboard.
System.NullReferenceException: Object reference not set to an instance of an object.
at uHelpsy.Helpers.ExamineIndexHelper.GetValueFromFieldOrProperty(IndexingNodeDataEventArgs e, String luceneFieldName, String propertyAlias) in d:\_PROJECTS\Personal\uHelpsy - Library\uHelpsy\Source\uHelpsy.Web\Helpers\ExamineIndexHelper.cs:line 98
Sorry for the delay (was at uWestFest). If you are getting thread aborts, that can cause IIS to restart the app pool, which would slow your site down. And I have also had issues with uCommentsy/uBlogsy slowing down page load times by a few seconds on the frontend of websites. It's probably worse for sites with more pages/comments.
I am not sure why you are doing a redirect and how that would avoid a thread abort exception. Can you explain a little about why you are doing that and why you think it would prevent a thread abort exception?
I am doing a redirect because there is a portion in the site where it pulls pages from a different database.
Its a different CMS completely (custom CMS and not umbraco). Long story short, data contents update is done internally, has a very complex schema and there is an approval process they go through.
Umbraco is getting slower and slower
Back office and the site itself is getting slower. this is the stats from fiddler:
And this is a "forgiving" time. there are times it takes 5 minutes or more.
(the same performance occurs on both staging and production environment) I already have done this on the staging environment to no avail. http://stackoverflow.com/questions/25225573/visual-studio-2012-mvc-umbraco-site-extremely-slow-while-debugging
We have to keep restarting the server. The thing is our dev environment is also really really slow.
The site is in a dedicated server (both staging and production). On the staging environment, the sql server database is taking up 2GB of RAM
Does anybody know is this normal?
Argh, the Umbraco forum ate my reply (as it sometimes does, annoyingly), so I'll reply again more briefly:
Definitely disable some of your dashboards then. In particular, I have had major problems with the uCommentsy dashboard before. Whenever I logged in, that dashboard would show up, and it would take a minute or two for the backoffice to load. It had some seriously slow code that I think may have been optimized a bit (I submitted a pull request, but I don't remember the details). Upgrading might help, but I'd definitely start by disabling the uCommentsy comments approval dashboard.
Also, no, you don't need Visual Studio for dotTrace. You can do a standalone install. You can attach it to a running process to do some profiling.
omg. disabling uCommentsy and uBlogsy did it! I am most grateful to you Nicholas!
thank you soo very much.
Now that the back office has improved, i need to check on the public facing site.
After two days, the back office is fast now. but the public facing site is still slow and i believe it is caused by UBlogsy.
I keep getting thread aborts and this is a partial view of umbraco Logs
2015-03-05 11:13:23,117 [154] WARN umbraco.macro - [Thread 245] Error loading MacroEngine script (file: SEOItems.cshtml, Type: ''. Exception: System.Threading.ThreadAbortException: Thread was being aborted.
at ASP._Page_macroScripts_SEOItems_cshtml.Execute() in d:\Inetpub\wwwprod\MacroScripts\SEOItems.cshtml:line 122
at System.Web.WebPages.WebPageBase.ExecutePageHierarchy()
at System.Web.WebPages.WebPage.ExecutePageHierarchy(IEnumerable`1 executors)
at System.Web.WebPages.WebPage.ExecutePageHierarchy()
at System.Web.WebPages.WebPageBase.ExecutePageHierarchy(WebPageContext pageContext, TextWriter writer, WebPageRenderingBase startPage)
at umbraco.MacroEngines.RazorMacroEngine.ExecuteRazor(MacroModel macro, INode currentPage)
at umbraco.MacroEngines.RazorMacroEngine.Execute(MacroModel macro, INode currentPage)
at umbraco.macro.loadMacroScript(MacroModel macro)
at umbraco.macro.renderMacro(Hashtable pageElements, Int32 pageId)
2015-03-05 11:13:26,836 [154] ERROR System.Exception - [Thread 230] During indexing luceneFieldName:uBlogsySearchableLabels, propertyAlias:uBlogsyPostLabels.
System.NullReferenceException: Object reference not set to an instance of an object.
at uHelpsy.Helpers.ExamineIndexHelper.GetValueFromFieldOrProperty(IndexingNodeDataEventArgs e, String luceneFieldName, String propertyAlias) in d:\_PROJECTS\Personal\uHelpsy - Library\uHelpsy\Source\uHelpsy.Web\Helpers\ExamineIndexHelper.cs:line 98
this is solved. the backoffice is fast now but the front end is still slow.
i am getting rid of a lot of ublogsy errors. hopefully it works.
Sorry for the delay (was at uWestFest). If you are getting thread aborts, that can cause IIS to restart the app pool, which would slow your site down. And I have also had issues with uCommentsy/uBlogsy slowing down page load times by a few seconds on the frontend of websites. It's probably worse for sites with more pages/comments.
hello nicholas.
So i did our redirects to use:
Response.Redirect("/page-not-found/", false);
Context.ApplicationInstance.CompleteRequest();
but it still triggers thread abort error. Is there another way of doing this?
~ i forced a 404 error and did a redirect to another page. it does an abort thread exception with the response.redirect.
I am not sure why you are doing a redirect and how that would avoid a thread abort exception. Can you explain a little about why you are doing that and why you think it would prevent a thread abort exception?
Happy New Year!
I am doing a redirect because there is a portion in the site where it pulls pages from a different database.
Its a different CMS completely (custom CMS and not umbraco). Long story short, data contents update is done internally, has a very complex schema and there is an approval process they go through.
the redirect issues were never solved.
is working on a reply...