Takes time to Save and publish

Question

Project server 2010 + SQL server 2005 sp3
When the users make the required changes and when they try to save and publish it takes them around 15 seconds to complete.
What would cause that to happen, and how can one improve the performance?
Satyam.

Answer 1

Let's try to look at it in a different way! When you save and publish a project there will be a series of job which will be triggered in the queue like Project Save from Project Professional, Project Publish, Reporting Project Publish etc.
Please let us know which among this is taking time.
[Please note some of the jobs gets pushed into the queue at the same time so I would suggest that you monitor the queue as soon as the save and publish is triggered and track the time taken for the jobs to complete]
Below I have also listed some of the recommended DB maintenance queries
1. Check size of Shadow Tables : http://support.microsoft.com/kb/2598007
  i. Run the below query on both the Draft and Published databases
select count(*) from MSP_TASK_CUSTOM_FIELD_VALUES_SHADOW 
select count(*) from MSP_ASSN_CUSTOM_FIELD_VALUES_SHADOW 
select count(*) from MSP_PROJ_RES_CUSTOM_FIELD_VALUES_SHADOW
select count(*) from MSP_PROJ_CUSTOM_FIELD_VALUES_SHADOW 
select count(*) from MSP_PROJ_CUSTOM_FIELD_VALUES_SHADOW 
  ii. Run the below query on Published database only
select count(*) from MSP_RES_CUSTOM_FIELD_VALUES_SHADOW
2. Update Statistics: Updates query optimization statistics on a table or indexed view. By default, the query optimizer already updates statistics as necessary to improve the query plan; in some cases you can improve query performance by using UPDATE STATISTICS
or the stored procedure sp_updatestats to update statistics more frequently than the default updates (References: http://technet.microsoft.com/en-us/library/ms187348.aspx, http://blogs.technet.com/b/projectadministration/archive/2009/12/04/sql-server-settings-for-performance-recommendations.aspx) 
Run the below query on all the four Project databases
sp_updatestats
3. Defragment all the four Project Server databases by running the below query on each of the databases to help improve the performance(This mentions Project 2007 bus is applicable for Project 2010 as well): http://support.microsoft.com/kb/943345
SET NOCOUNT ON
    DECLARE @objectid int
    DECLARE @indexid int
    DECLARE @command varchar(8000)
    DECLARE @baseCommand varchar(8000)
    DECLARE @schemaname sysname
    DECLARE @objectname sysname
    DECLARE @indexname sysname
    DECLARE @currentDdbId int
    SELECT @currentDdbId = DB_ID()
    PRINT CONVERT(nvarchar, GETDATE(), 126) + ': Starting'
    -- Loop over each of the indices
    DECLARE indexesToDefrag CURSOR FOR 
    SELECT 
        i.object_id, 
        i.index_id, 
        i.name
    FROM 
        sys.indexes AS i
    INNER JOIN 
        sys.objects AS o
    ON
        i.object_id = o.object_id
    WHERE 
        i.index_id > 0 AND
        o.type = 'U'
    OPEN indexesToDefrag
    -- Loop through the partitions.
    FETCH NEXT
    FROM
        indexesToDefrag
    INTO 
        @objectid, 
        @indexid,
        @indexname
    WHILE @@FETCH_STATUS = 0
    BEGIN
        -- Lookup the name of the index
        SELECT 
            @schemaname = s.name
        FROM 
            sys.objects AS o
        JOIN 
            sys.schemas AS s
        ON
            s.schema_id = o.schema_id
        WHERE
            o.object_id = @objectid
        PRINT CONVERT(nvarchar, GETDATE(), 126) + ': ' + @schemaname + '.' + @indexname + ' is now being rebuilt.'
        -- Fragmentation is bad enough that it will be more efficient to rebuild the index
        SELECT @baseCommand = 
            ' ALTER INDEX ' + 
                @indexname +
            ' ON ' + 
                @schemaname + '.' + object_name(@objectid) + 
            ' REBUILD WITH (FILLFACTOR = 80, ONLINE = '
        -- Use dynamic sql so this compiles in SQL 2000
        SELECT @command =
            ' BEGIN TRY ' + 
               @baseCommand + 'ON) ' +
            ' END TRY ' +
            ' BEGIN CATCH ' +
               -- Indices with image-like columns can't be rebuild online, so go offline
               @baseCommand + 'OFF) ' +
            ' END CATCH '
        PRINT CONVERT(nvarchar, GETDATE(), 126) + ': Rebuilding'
        EXEC (@command)
        PRINT CONVERT(nvarchar, GETDATE(), 126) + ': Done'
        FETCH NEXT FROM indexesToDefrag INTO @objectid, @indexid, @indexname
    END
    CLOSE indexesToDefrag
    DEALLOCATE indexesToDefrag
4. Recompile Stored Procedures: Empty the stored execution plans for the below stored procedures from the cache and then recompile them so that the new execution plan get things working faster (Reference# http://blogs.msdn.com/b/brismith/archive/2012/09/19/when-your-project-server-queue-slows-down.aspx)
Draft DB 
EXEC Sp_recompile MSP_ProjQ_Lock_Next_Available_Group 
EXEC Sp_recompile MSP_ProjQ_Get_Status_Of_Jobs_List 
Published DB 
EXEC Sp_recompile MSP_TimesheetQ_Get_Status_Of_Jobs_List 
EXEC Sp_recompile MSP_TimesheetQ_Lock_Next_Available_Group
Cheers! Happy troubleshooting !!! Dinesh S. Rai - MSFT Enterprise Project Management Please click Mark As Answer; if a post solves your problem or Vote As Helpful if a post has been useful to you. This can be beneficial to other community members reading
the thread.