Optimize the LINQ to Entities group
I have this query in LINQ to Entities.
var query = (from s in db.ForumStatsSet
where s.LogDate >= date1 && s.LogDate <= date2
group s by new { s.Topic.topicID, s.Topic.subject, s.Topic.Forum.forumName, s.Topic.datum, s.Topic.Forum.ForumGroup.name, s.Topic.Forum.forumID } into g
orderby g.Count() descending
select new TopicStatsData
{
TopicId = g.Key.topicID,
Count = g.Count(),
Subject = g.Key.subject,
ForumGroupName = g.Key.name,
ForumName = g.Key.forumName,
ForumId = g.Key.forumID
});
I know this is a kind of "evil" request, but it is only used in the admin interface. But the SQL he created is absolutely terrifying. Look at this child.
exec sp_executesql N'SELECT TOP (50)
[Project6].[C1] AS [C1],
[Project6].[TopicId] AS [TopicId],
[Project6].[C4] AS [C2],
[Project6].[subject] AS [subject],
[Project6].[name] AS [name],
[Project6].[forumName] AS [forumName],
[Project6].[C2] AS [C3]
FROM ( SELECT
[Project5].[TopicId] AS [TopicId],
[Project5].[subject] AS [subject],
[Project5].[forumName] AS [forumName],
[Project5].[name] AS [name],
1 AS [C1],
CAST( [Project5].[forumID] AS int) AS [C2],
[Project5].[C1] AS [C3],
[Project5].[C2] AS [C4]
FROM ( SELECT
[Project4].[TopicId] AS [TopicId],
[Project4].[forumID] AS [forumID],
[Project4].[subject] AS [subject],
[Project4].[forumName] AS [forumName],
[Project4].[name] AS [name],
[Project4].[C1] AS [C1],
(SELECT
COUNT(cast(1 as bit)) AS [A1]
FROM [dbo].[tForumStats] AS [Extent14]
LEFT OUTER JOIN [dbo].[tTopic] AS [Extent15] ON [Extent14].[TopicId] = [Extent15].[topicID]
LEFT OUTER JOIN [dbo].[tForum] AS [Extent16] ON [Extent15].[forumID] = [Extent16].[forumID]
LEFT OUTER JOIN [dbo].[tForum] AS [Extent17] ON [Extent15].[forumID] = [Extent17].[forumID]
LEFT OUTER JOIN [dbo].[tForum] AS [Extent18] ON [Extent15].[forumID] = [Extent18].[forumID]
LEFT OUTER JOIN [dbo].[tForumGroup] AS [Extent19] ON [Extent18].[forumGroupID] = [Extent19].[forumGroupID]
LEFT OUTER JOIN [dbo].[tForum] AS [Extent20] ON [Extent15].[forumID] = [Extent20].[forumID]
LEFT OUTER JOIN [dbo].[tForumGroup] AS [Extent21] ON [Extent20].[forumGroupID] = [Extent21].[forumGroupID]
WHERE ([Extent14].[LogDate] >= @p__linq__25) AND ([Extent14].[LogDate] = @p__linq__25) AND ([Extent6].[LogDate] = @p__linq__25) AND ([Extent1].[LogDate]
I do not as anyone to explain that query but it would be great to get some tips on how to optimze the query so that it just do a simple regular join. Something like this works as fine if I write the SQL myself.
SELECT COUNT (*) AS NumberOfViews, s.topicid AS topicId, t.subject AS TopicSubject, g. [Name] AS ForumGroupName, f.forumName AS ForumName FROM tForumStats s join tTopic t on s.topicid = t.topicid join tForum f on f.forumid = t.forumid JOIN tForumGroup g ON f.forumGroupID = g.forumGroupID WHERE s. [LogDate] between @ date1 AND @ date2 group by s.topicid, t.subject, f.Forumname, t.Datum, g. [name] order by count (*) desc
Btw, I LOVE this site. Amazing design and usability! Hope it works well to help :)
+1
source to share
2 answers
Instead of joining all tables in a group, you can join these tables yourself. Can you try this;
from s in db.ForumStatsSet
join t in db.Topics on t.TopicId == s.TopicId
join f in db.Forums on f.ForumId == t.ForumId
join fg in db.ForumGroups on fg.ForumGroupId == f.ForumGroupId
where s.LogDate >= date1 && s.LogDate <=
group s by new { t.TopicId, t.subject, f.forumName, t.datum, fg.name, f.forumID } into g
orderby g.Count() descending
select new TopicStatsData
{
TopicId = g.Key.topicID,
Count = g.Count(),
Subject = g.Key.subject,
ForumGroupName = g.Key.name,
ForumName = g.Key.forumName,
ForumId = g.Key.forumID
});
ps: there may be some errors, but logically it should be correct!
+5
source to share