I have the following code which updates / adds data to a SQL Server CE database with EF6, which is working fine for small number of records. However when the volume of records exceeds 1000~2000 the transaction become very slow (10~15sec). Is there any way to optimize it?
Database.DefaultConnectionFactory = new SqlCeConnectionFactory("System.Data.SqlServerCe.4.0", "", MyProject.ConnectionString);
ProjectContext context = new ProjectContext();
context.Configuration.AutoDetectChangesEnabled = false;
context.Configuration.ValidateOnSaveEnabled = false;
using (var db = new ProjectContext())
{
foreach (var item in MyProject.Brands)
{
if (!db.Brands.Any(i => i.Name == item.Name))
{
// Add
db.Brands.Add(item);
}
else
{
// Update
var found = db.Brands.First(i => i.Name == item.Name);
found = item;
}
}
db.SaveChangesAsync();
This way you save 50% on your Linq queries, and your update might actually work (your MyProject.Brands are not attached in any way to your new ProjectContext that you are using, the found that you extract from there is attached, but in your original code, you overwrote it with yuor item, meaning it will be ignored.
context I have left out completely, as you were not doing anything with it, as your using creates a new ProjectContext anyway.
using (var db = new ProjectContext())
{
foreach (var item in MyProject.Brands)
{
var found = db.Brands.FirstOrdefault(i => i.Name == item.Name);
if (found == null)
{
// Add
db.Brands.Add(item);
}
else
{
// Update
found.prop1 = item.prop1;
found.prop2 = item.prop2;
// ... etc, for all and any updatable properties
}
}
db.SaveChangesAsync();
}
Related
So i want to update an entry's Valid to column, and then insert a new copy of it with a new value. The issue is it seems to skip the update statement, and just inserts a new copy of it.
foreach (Model data in Data)
{
var entry = context.table.Where(x=>x.id == data.id).FirstOrDefault();
entry.ValidTo = DateTime.Now;
ctx.Update(entry);
entry.id = 0;
entry.ValidTo = new DateTime(9999, 12, 31);
entry.ValidFrom = DateTime.Now;
entry.Value = adjustmentmodel.Value;
ctx.Add(entry);
}
ctx.SaveChanges();
I tried inserting a saveChanges after ctx.update(entry), and that works, but is pretty slow. So i wanted to hear if there was a way, to only have to save the changes at the end?
I am using dotnet 5 and EF core 5.0.17
Separate your entity references, there is no reason to re-use it.
foreach (Model data in Data)
{
// Update the entity
var entry = context.table.Where(x => x.id == data.id).FirstOrDefault();
entry.ValidTo = DateTime.Now;
// Add a new entry
ctx.Add(new Entry
{
// what you need for the new entry
});
}
// Save the changes within a single transaction
ctx.SaveChanges();
Please try using UpdateRange and AddRange method once.
var entry = Context.table.Where(x => Data.Select(y => y.id).Contains(x.id)).ToList();
entry = entry.ForEach(res =>
{
res.ValidTo = DateTime.Now
}).ToList();
ctx.UpdateRange(entry);
entry = entry.ForEach(res =>
{
res.ValidTo = new DateTime(9999, 12, 31);
res.ValidFrom = DateTime.Now;
res.Value = adjustmentmodel.Value;
}).ToList();
ctx.AddRange(entry);
ctx.SaveChanges();
IEnumerable<WebsiteWebPage> data = GetWebPages();
foreach (var value in data)
{
if (value.WebPage.Contains(".htm"))
{
WebsiteWebPage pagesinfo = new WebsiteWebPage();
pagesinfo.WebPage = value.WebPage;
pagesinfo.WebsiteId = websiteid;
db.WebsiteWebPages.Add(pagesinfo);
}
}
db.SaveChanges();
I want to add only distinct values to database in above code. Kindly help me how to do it as I am not able to find any solution.
IEnumerable<WebsiteWebPage> data = GetWebPages();
foreach (var value in data)
{
if (value.WebPage.Contains(".htm"))
{
var a = db.WebsiteWebPages.Where(i => i.WebPage == value.WebPage.ToString()).ToList();
if (a.Count == 0)
{
WebsiteWebPage pagesinfo = new WebsiteWebPage();
pagesinfo.WebPage = value.WebPage;
pagesinfo.WebsiteId = websiteid;
db.WebsiteWebPages.Add(pagesinfo);
db.SaveChanges();
}
}
}
This is the code that I used to add distinct data.I hope it helps
In addition to the code sample Furkan Öztürk supplied, Make sure your DB has a constraint so that you cannot enter duplicate values in the column. Belt and braces approach.
I assume that by "distinct values" you mean "distinct value.WebPage values":
// get existing values (if you ever need this)
var existingWebPages = db.WebsiteWebPages.Select(v => v.WebPage);
// get your pages
var webPages = GetWebPages().Where(v => v.WebPage.Contains(".htm"));
// get distinct WebPage values except existing ones
var distinctWebPages = webPages.Select(v => v.WebPage).Distinct().Except(existingWebPages);
// create WebsiteWebPage objects
var websiteWebPages = distinctWebPages.Select(v =>
new WebsiteWebPage { WebPage = v, WebsiteId = websiteid});
// save all at once
db.WebsiteWebPages.AddRange(websiteWebPages);
db.SaveChanges();
Assuming that you need them to be unique by WebPage and WebSiteId
IEnumerable<WebsiteWebPage> data = GetWebPages();
foreach (var value in data)
{
if (value.WebPage.Contains(".htm"))
{
WebsiteWebPage pagesinfo = new WebsiteWebPage();
if (db.WebsiteWebPages.All(c=>c.WebPage != value.WebPage|| c.WebsiteId != websiteid))
{
pagesinfo.WebPage = value.WebPage;
pagesinfo.WebsiteId = websiteid;
db.WebsiteWebPages.Add(pagesinfo);
}
}
}
db.SaveChanges();
UPDATE
To optimize this (given that your table contains much more data than your current list), override your equals in WebsiteWebPage class to define your uniqueness criteria then:
var myWebsiteWebPages = data.select(x=> new WebsiteWebPage { WebPage = x.WebPage, WebsiteId = websiteid}).Distinct();
var duplicates = db.WebsiteWebPages.Where(x=> myWebsiteWebPage.Contains(x));
db.WebsiteWebPages.AddRange(myWebsiteWebPages.Where(x=> !duplicates.Contains(x)));
this is a one database query to retrieve ONLY duplicates and then removing them from the list
You can use the following code,
IEnumerable<WebsiteWebPage> data = GetWebPages();
var templist = new List<WebsiteWebPage>();
foreach (var value in data)
{
if (value.WebPage.Contains(".htm"))
{
WebsiteWebPage pagesinfo = new WebsiteWebPage();
pagesinfo.WebPage = value.WebPage;
pagesinfo.WebsiteId = websiteid;
templist.Add(pagesinfo);
}
}
var distinctList = templist.GroupBy(x => x.WebsiteId).Select(group => group.First()).ToList();
db.WebsiteWebPages.AddRange(distinctList);
db.SaveChanges();
Or you can use MoreLINQ here to filter distinct the list by parameter like,
var res = tempList.Distinct(x=>x.WebsiteId).ToList();
db.WebsiteWebPages.AddRange(res);
db.SaveChanges();
What is the best way to update multiple records in a list to speed up processing?
Currently, I'm updating about 15000 products, each with 3 different price sets and it takes the whole day to complete.
I need to update the prices all at once in code side, then commit those changes to the database in 1 go, instead of fetching each inventory item, updating its values, then attaching it to the context. Every single fetch is causing the delays.
Code
public void UpdatePricesFromInventoryList(IList<Domain.Tables.Inventory> invList)
{
var db = new UniStockContext();
foreach (var inventory in invList)
{
Domain.Tables.Inventory _inventory = db.Inventories
.Where(x => x.InventoryID == inventory.InventoryID)
.FirstOrDefault();
if (inventory.Cost.HasValue)
_inventory.Cost = inventory.Cost.Value;
else
_inventory.Cost = 0;
foreach (var inventoryPrices in inventory.AccInventoryPrices)
{
foreach (var _inventoryPrices in _inventory.AccInventoryPrices)
{
if (_inventoryPrices.AccInventoryPriceID == inventoryPrices.AccInventoryPriceID)
{
_inventoryPrices.ApplyDiscount = inventoryPrices.ApplyDiscount;
_inventoryPrices.ApplyMarkup = inventoryPrices.ApplyMarkup;
if (inventoryPrices.Price.HasValue)
_inventoryPrices.Price = inventoryPrices.Price.Value;
else
_inventoryPrices.Price = _inventory.Cost;
if (inventoryPrices.OldPrice.HasValue)
{
_inventoryPrices.OldPrice = inventoryPrices.OldPrice;
}
}
}
}
db.Inventories.Attach(_inventory);
db.Entry(_inventory).State = System.Data.Entity.EntityState.Modified;
}
db.SaveChanges();
db.Dispose();
}
I've also tried working my code according to this SOQ Entity Framework update/insert multiple entities
and it gave me and error. Here are the details:
Code:
public void UpdatePricesFromInventoryListBulk(IList<Domain.Tables.Inventory> invList)
{
var accounts = new List<Domain.Tables.Inventory>();
var db = new UniStockContext();
db.Configuration.AutoDetectChangesEnabled = false;
foreach (var inventory in invList)
{
accounts.Add(inventory);
if (accounts.Count % 1000 == 0)
{
db.Set<Domain.Tables.Inventory>().AddRange(accounts);
accounts = new List<Domain.Tables.Inventory>();
db.ChangeTracker.DetectChanges();
db.SaveChanges();
db.Dispose();
db = new UniStockContext();
}
}
db.Set<Domain.Tables.Inventory>().AddRange(accounts);
db.ChangeTracker.DetectChanges();
db.SaveChanges();
db.Dispose();
}
Error:
An entity object cannot be referenced by multiple instances of IEntityChangeTracker.
I would suggest changing the following:
Domain.Tables.Inventory _inventory = db.Inventories
.Where(x => x.InventoryID == inventory.InventoryID)
.FirstOrDefault();
To
Domain.Tables.Inventory _inventory = db.Inventories
.Single(x => x.InventoryID == inventory.InventoryID);
I'd still add the db.Configuration.AutoDetectChangesEnabled = false; after getting the context, and also use AsNoTracking:
Turn off EF change tracking for any instance of the context
that is because you are hit the database context at every loop to increase the performance you should get all the Inventories by one hit ,this is your problem try the below code and you will notice the performance :
public void UpdatePricesFromInventoryList(IList<Domain.Tables.Inventory> invList)
{
var db = new UniStockContext();
invIdsArray = invList.select(x => x.InventoryID).ToArray();
IList<Domain.Tables.Inventory> invListFromDbByOneHit = db.Inventories.Where(x => invIdsArray.Contains(x.InventoryID)).Tolist();
foreach (var inventory in invListFromDbByOneHit)
{
//Domain.Tables.Inventory _inventory = db.Inventories
//.Where(x => x.InventoryID == inventory.InventoryID)
//.FirstOrDefault();
if (inventory.Cost.HasValue)
_inventory.Cost = inventory.Cost.Value;
else
_inventory.Cost = 0;
foreach (var inventoryPrices in inventory.AccInventoryPrices)
{
foreach (var _inventoryPrices in _inventory.AccInventoryPrices)
{
if (_inventoryPrices.AccInventoryPriceID == inventoryPrices.AccInventoryPriceID)
{
_inventoryPrices.ApplyDiscount = inventoryPrices.ApplyDiscount;
_inventoryPrices.ApplyMarkup = inventoryPrices.ApplyMarkup;
if (inventoryPrices.Price.HasValue)
_inventoryPrices.Price = inventoryPrices.Price.Value;
else
_inventoryPrices.Price = _inventory.Cost;
if (inventoryPrices.OldPrice.HasValue)
{
_inventoryPrices.OldPrice = inventoryPrices.OldPrice;
}
}
}
}
db.Inventories.Attach(_inventory);
db.Entry(_inventory).State = System.Data.Entity.EntityState.Modified;
}
db.SaveChanges();
db.Dispose();
}
I pull student data from 2 databases. 1 from an online SOAP API which can handle async calls and 1 from a local DB with an older services that doesnt support async.
I compare these databases and write the differences in a local sqlDB through EF.
Problem:
I get double entries in my EF DB. He puts the correct data and amount in arrays inside the method, but it looks like once he hits the db.savechanges() he jumps back up a few line and saves again.
I don't even know where this extra thread comes from.
Some code might be still there from numerous tries to solve it. For instance I tried with addrange but I get an error when he tries to add the FullVarianceList.
public async Task<bool> FullStudentCompare(string date) //format DD/MM/YYYY
{
try
{
//DB context
using (var db = new SchoolDbContext())
{
//GET DATA
//SMT (async)
List<SmtStudent> smtStdudentList = await GetAllSmartschoolStudents();
//Wisa (sync)
//on date, or if emty on current systemdate
List<WisaStudent> wisaList;
if (date == "")
{
wisaList = GetWisaStudentData(DateTime.Now.ToShortDateString());
}
else
{
wisaList = GetWisaStudentData(date);
}
//Flags and props needed for DB entry after compare
bool existsInLocalDb = false;
List<Variance> vList = new List<Variance>();
//Full list to add to DB outside foreach
List<Variance> fullVarianceList = new List<Variance>();
//Full List of new Students to write to DB outside foreach
List<DbStudent> fullStudentList = new List<DbStudent>();
//Compare lists
foreach (WisaStudent wstd in wisaList)
{
//determine correct classCode
string klasCode;
if (wstd.klasgroep.Trim() == "Klasgroep 00")
{
klasCode = wstd.klas.Trim();
}
else
{
klasCode = wstd.klasgroep.Trim();
}
//Create SmtStudent object for compare
SmtStudent tempStd = new SmtStudent(true,
wstd.voornaam.Trim(),
wstd.naam.Trim(),
wstd.stamboeknummer.Trim(),
wstd.geslacht.Trim(),
wstd.geboortedatum.Trim(),
wstd.straat.Trim(),
wstd.huisnummer.Trim(),
wstd.busnummer.Trim(),
wstd.postcode.Trim(),
wstd.gemeente.Trim(),
wstd.emailadres.Trim(),
wstd.GSM_nummer.Trim(),
wstd.levensbeschouwing.Trim(),
wstd.coaccountmoedervoornaam.Trim(),
wstd.coaccountmoedernaam.Trim(),
wstd.coaccountmoederemailadres.Trim(),
wstd.coaccountmoederGSM_nummer.Trim(),
wstd.coaccountvadervoornaam.Trim(),
wstd.coaccountvadernaam.Trim(),
wstd.coaccountvaderemailadres.Trim(),
wstd.coaccountvaderGSM_nummer.Trim(),
klasCode,
wstd.nationaliteit,
wstd.geboorteGemeente,
wstd.geboorteLand
);
//Find matching SmtStudent
SmtStudent smtStd = smtStdudentList.Find(i => i.Internnummer == wstd.stamboeknummer);
//Find matching Std in local DB
DbStudent dbStd = await db.Students.Where(i => i.Stamboeknummer == wstd.stamboeknummer).FirstOrDefaultAsync();
//if none exists in the local DB create an entity to update and write to DB
if (dbStd == null)
{
dbStd = new DbStudent(wstd.voornaam.Trim(),
wstd.naam.Trim(),
wstd.stamboeknummer.Trim(),
wstd.geslacht.Trim(),
wstd.geboortedatum.Trim(),
wstd.straat.Trim(),
wstd.huisnummer.Trim(),
wstd.busnummer.Trim(),
wstd.postcode.Trim(),
wstd.gemeente.Trim(),
wstd.emailadres.Trim(),
wstd.GSM_nummer.Trim(),
wstd.levensbeschouwing.Trim(),
wstd.coaccountmoedervoornaam.Trim(),
wstd.coaccountmoedernaam.Trim(),
wstd.coaccountmoederemailadres.Trim(),
wstd.coaccountmoederGSM_nummer.Trim(),
wstd.coaccountvadervoornaam.Trim(),
wstd.coaccountvadernaam.Trim(),
wstd.coaccountvaderemailadres.Trim(),
wstd.coaccountvaderGSM_nummer.Trim(),
klasCode,
wstd.loopbaanDatum,
wstd.nationaliteit,
wstd.geboorteGemeente,
wstd.geboorteLand
);
db.Students.Add(dbStd);
fullStudentList.Add(dbStd);
}
else
{
existsInLocalDb = true;
}
if (smtStd == null)
{
//Std doesn't exist in Smt -> New student
dbStd.IsNewStudent = true;
dbStd.ClassMovement = true;
//remove from wisaList
wisaList.Remove(wstd);
}
else
{
//clear vlist from previous iterations
vList.Clear();
//get all properties on the obj, cycle through them and find differences
PropertyInfo[] props = smtStd.GetType().GetProperties();
vList.AddRange(props.Select(f => new Variance
{
Property = f.Name,
ValueA = f.GetValue(smtStd),
ValueB = f.GetValue(tempStd),
Checked = false
})
.Where(v => !v.ValueA.Equals(v.ValueB) && v.ValueB != null)
.ToList());
//If the users allrdy exists in LocalDb delete all previously recorded variances
if (existsInLocalDb)
{
if (db.Variances.Where(j => j.Student.StudentId.Equals(dbStd.StudentId)).FirstOrDefault() != null)
{ //if the student allready exists we will recreate the variancelist, hence deleting all current items first
List<Variance> existingList = db.Variances.Where(j => j.Student.StudentId.Equals(dbStd.StudentId)).ToList();
foreach (Variance v in existingList)
{
db.Variances.Remove(v);
}
}
}
//Add new variances if vList is not empty
if (vList.Count > 0)
{
//Check if KlasCode is a variance -> set classmovement to true
if (vList.Where(i => i.Property == "KlasCode").FirstOrDefault() != null)
{
dbStd.ClassMovement = true;
}
else
{
dbStd.ClassMovement = false;
}
//add the StudentObject to the variance to link them 1-many
foreach (Variance v in vList)
{
v.Student = dbStd;
fullVarianceList.Add(v);
db.Variances.Add(v);
}
}
}
}
//add the full lists of variances and new students to DB
//db.Variances.AddRange(fullVarianceList);
//db.Students.AddRange(fullStudentList);
db.SaveChanges();
return true;
}
}
catch(Exception ex)
{
return false;
}
}
A couple of things:
It is important to understand that EF uses a unit of work pattern where none of the changes to the entities are persisted until SaveChanges is called which explains the "once he hits the db.Savechanges() he jumps back up" phenomenon.
When you have a 1 to many relationsship and you assign a collection of entities to a navigation property on another entity and then add that parent entity to the DbContext, EF marks those child entities to be added too. In your case dbStd is added at the line "db.Students.Add(dbStd);" and at the line "v.Student = dbStd;". This is most likely what is causing your duplicates.
I have for example database with codes entities and I want in single transaction read and update some information.
What will be marked and "locked" or "dirty" when I execute such statement from Entity Framework 6.
using(var context = new codesEntities())
{
using (var trans = db.Database.BeginTransaction(System.Data.IsolationLevel.Snapshot))
{
var codes_in_group = context.codes.Where(x => x.group == 1);
}
}
I am not sure but maybe nothing since nothing is read from querable.
Is it anything marked as read?
What will be marked as read if execute such code?
using(var context = new codesEntities())
{
using (var trans = db.Database.BeginTransaction(System.Data.IsolationLevel.Snapshot))
{
var codes_in_group = context.codes.Where(x => x.group == 1).ToList();
}
}
Or such code:
using(var context = new codesEntities())
{
using (var trans = db.Database.BeginTransaction(System.Data.IsolationLevel.Snapshot))
{
var codes_in_group = context.codes.Where(x => x.group == 1).GetEnumerator();
}
}