将数据上传到SQL表（实体框架）后，实体按时间错误的方式排序

Question

I have a problem with Id order in my SQL table. I created a program, that reads some variables from machine and then records these values into SQL table by means of Entity Framework. For uploading of values it uses two lists - when the first list is filled (time of filling, for example, is about 10 minutes) with variables, the second list (that contains recorded variables) uploads to SQL table and clears itself. After 10 minutes lists change.

It works fine, but once I noticed that some rows (some = 100 rows from 10000 rows) have "broken order". I mean there is something like below:

ID - Value - Time

1000 - 3 - 15:00

1001 - 4 - 15:01

1002 - 1 - 14:58 !!

1003 - 2 - 14:59 !!

1004 - 5 - 15:02

1005 - 6 - 15:03

In other words there are some variables that sometimes change their order in SQL table. Thousands rows before and after these problem rows are OK.

UPD: As I just find out, this problem occurs every 50 th and 51 th minutes if time (14:50:xx...15:50)... very strange for understanding...

So I tried to edit code a little (don´t throw bricks at me - I am quite new in programming). There is a method that writes recorded variables from lists to SQL table:

public void WriteRecordedVariableToSQL()
{
    if (timeTumbler == true && VariablesToBulkSQLList2.Count > 0)
    {
        List<VariableRecord> sortedVariablesToBulkSQLList2 = VariablesToBulkSQLList2.OrderBy(o => o.Time).ToList();
        context.VariableRecords.AddRange(sortedVariablesToBulkSQLList2);
        context.BulkSaveChanges();
        VariablesToBulkSQLList2.Clear();
        sortedVariablesToBulkSQLList2.Clear();

    }

    if (timeTumbler == false && VariablesToBulkSQLList1.Count > 0)
    {
        List<VariableRecord> sortedVariablesToBulkSQLList1 = VariablesToBulkSQLList1.OrderBy(o => o.Time).ToList(); //sortuju podle casu
        context.VariableRecords.AddRange(sortedVariablesToBulkSQLList1);
        context.BulkSaveChanges();
        VariablesToBulkSQLList1.Clear();
        sortedVariablesToBulkSQLList1.Clear();

    }

}

But the result was the same. As you see, I use Entity FrameWork Extentions( EFE ) and their AddRange for adding the entire lists. May the problem be there?

So I´d like to fix this problem that there are values ordered by time in SQL table (with right order of ID).

Can you help me? Thanks in advance. Hope you understand my poor English.

Answer 1

Here's a little on the basics. Id in your case (a generated PK by SQL Server) is a meaningless field. In other words, it is a field that conveys no useful information beyond being a unique key for that particular row in the table. It does not specify order or sequence.

For your example, sorting is part of the data retrieval and would involve selecting (and sorting) based upon the field you use to store the time. It does not affect or influence the primary key.

Add-on: It will hurt you in the long run to attempt to "build in" intelligence with your primary keys. That is one reason why they are also referred to as "surrogate" keys. It is a piece of artificial information that is unchanging and refers to a unique row.

Other key types include "candidate" keys (unique but changeable) and "foreign" keys (primary key to another entity). Indices which are non-essential but useful for performance include "compound" keys which combine multiple fields (as a candidate key or for ease of sorting) and single field indices which are mostly for DB performance.