Hi Parvez,

100000 is a lot of blog items. In this the API using LINQ is not the best option performance wise. But there are other options like XpathNavigator or Examine.

Using a XpathNavigator you can do the following :

var pageSize = 10;
var currentPage = 2;
var itemsToSkip = (currentPage -1) * pageSize;

// very greedy xpath, can be made more specific for better performance
var xpath = "//BlogDocTypeAlias[@isDoc]";

var navigator = UmbracoContext.ContentCache.GetXPathNavigator();
var expression = navigator.Compile(xpath);
expression.AddSort("blogDateProp", XmlSortOrder.Descending, XmlCaseOrder.None, "", XmlDataType.Text);

var iterator = navigator.Select(expression).Cast<XPathNavigator>().Skip(itemsToSkip).Take(pageSize).ToList();

var blogs = new List<IPublisedContent>();

foreach(var item in iterator) 
{
    var contentId = int.Parse(item.GetAttribute("id",""));

    blogs.Add(UmbracoContext.ContentCache.GetById(id));
}

This works directly on the Umbraco XML cache. It will look up all blog items, sort them by date property, and take just the ones you need without loading them all in to memory like with linq

Another option is to use Examine

var pageSize = 10;
var currentPage = 2;
var itemsToSkip = (currentPage -1) * pageSize;
var blogs = new List<IPublisedContent>();

var searchCriteria = ExamineManager.Instance.DefaultSearchProvider.CreateSearchCriteria(IndexTypes.Content);

var query = searchCriteria.Field("nodeTypeAlias","BlogDocTypeAlias");

var results = ExamineManager.Instance.DefaultSearchProvider.Search(query.Compile())
                .OrderByDescending(x => x.Fields["blogDate"]).Skip(itemsToSkip).Take(pageSize);

foreach(var item in results) 
{
    blogs.Add(UmbracoContext.ContentCache.GetById(item.Id));
}

Both options will be a lot faster than a Linq query.

More info can be found here :

https://www.slideshare.net/dawoe/the-need-for-speed-uk-fest

Dave

is working on a reply...