There is still some room for improvement on network level. The packets are not
batches as much as it could be for server to still understand it. Dmitri Y.
reported it to me some months ago.
It just needs a lot of time with Wireshark and tweaking the socket until it's
same as with fbclient.dll.
Maybe somebody could help me with that as well.
--
Mgr. Jiří Činčura
Independent IT Specialist
________________________________
From: Alexander Muylaert-Gelein <amuylaert_gel...@hotmail.com>
Sent: Friday, August 08, 2014 11:10:43 PM
To: firebird-net-provider@lists.sourceforge.
Subject: Re: [Firebird-net-provider] Questions with performance metrics doing
large inserts into DB
Hi
Attached some examples of transactions properly set.
If you can put the reader in "readonly" and the writer in "WriteNoUndo", you
gain some speed. But euh, no magic to expect here. It might be interesting to
profile it though and post the results back.
public static class Transaction
{
private static FbTransactionOptions s_WriteTransaction = new
FbTransactionOptions()
{
TransactionBehavior = FbTransactionBehavior.Write |
FbTransactionBehavior.ReadCommitted |
FbTransactionBehavior.RecVersion |
FbTransactionBehavior.Wait,
WaitTimeout = new TimeSpan(0, 0, 0, 10) //seconds
};
public static FbTransactionOptions WriteTransactions
{
get
{
return s_WriteTransaction;
}
}
private static FbTransactionOptions s_WriteNoUndoTransaction = new
FbTransactionOptions()
{
TransactionBehavior = FbTransactionBehavior.Write |
FbTransactionBehavior.ReadCommitted |
FbTransactionBehavior.RecVersion |
FbTransactionBehavior.Wait |
FbTransactionBehavior.NoAutoUndo,
WaitTimeout = new TimeSpan(0, 0, 0, 10) //seconds
};
public static FbTransactionOptions WriteNoUndoTransactions
{
get
{
return s_WriteTransaction;
}
}
private static FbTransactionOptions s_ReadTransaction = new
FbTransactionOptions()
{
TransactionBehavior = FbTransactionBehavior.Read |
FbTransactionBehavior.ReadCommitted |
FbTransactionBehavior.RecVersion |
FbTransactionBehavior.NoWait
};
public static FbTransactionOptions ReadTransactions
{
get
{
return s_ReadTransaction;
}
}
public static FbTransaction BeginReadTransaction(this FbConnection
aConnection)
{
return aConnection.BeginTransaction(s_ReadTransaction);
}
public static FbTransaction BeginWriteTransaction(this FbConnection
aConnection)
{
return aConnection.BeginTransaction(s_WriteTransaction);
}
public static FbTransaction BeginWriteNoUndoTransaction(this FbConnection
aConnection)
{
return aConnection.BeginTransaction(s_WriteTransaction);
}
}
I read somewhere that Looping in .NET is somewhat slower than in Delphi, but
there are things you can do optimize the .NET loops.
Agreed, some tricks might be applied. But that will bring you down from 18.7
seconds to 18.69998 seconds. Optimizing loops is actually not the way to go.
What you could do is make a small test project. with a few simple tables. Add
your simulations to it and make sure it compiles. Then, if you have these
minimal projects building, I would like to receive a copy and I can profile for
you.
Maybe I can already see what is wrong. Like Jiri mentioned. .Net is slower
then this FIB/FBClient (c++) dll. But let us assume that is 10% overhead. So
4 seconds delphi => 5 seconds .net is fine for me. 18 seconds... I'm
interested in speeding things up then :-)
thanks
a
________________________________
From: emendez...@nc.rr.com
To: firebird-net-provider@lists.sourceforge.net
Date: Fri, 8 Aug 2014 09:41:03 -0400
Subject: Re: [Firebird-net-provider] Questions with performance metrics doing
large inserts into DB
Alexander,
Thank you for the feedback.
>>Two more things...
>>
>>1. Your transaction parameters please. How do you create them, what
>>settings?
To be honest I’ve never really thought about the transaction settings other
than the default. Are there a specific settings I should be using for just
reading “stale” data?
>>2. Did you check your "source-read" logic. Cound you maybe fake data, so we
>>know if it is the "insert" that is slow. Maybe the materializing of the
>>source record is slow or fetching it?
I was doing further testing last night and removed the insert logic to see if
it was the reading of the data that was slowing it down. with the removal of
the Insert Logic and just reading and looping for 5000 iterations, it still
slower than Delphi’s results. For the read Logic I was using Dapper dot net and
retrieving the results un-buffered (one row at a time). I then eliminated
Dapper and it still was slower than Delphi. I then tried an OleDB Provider and
that was a little faster than the .Net provider, but still slower than Delphi.
In retrospect I might have jumped the gun in blaming the Write performance of
the .Net Provider. I think I need to try to optimize the read logic, then move
on to the Write Logic.
I read somewhere that Looping in .NET is somewhat slower than in Delphi, but
there are things you can do optimize the .NET loops.
Thank you,
Edward Mendez
From: Alexander Muylaert-Gelein [mailto:amuylaert_gel...@hotmail.com]
Sent: Friday, August 8, 2014 1:54 AM
To: firebird-net-provider@lists.sourceforge.
Subject: Re: [Firebird-net-provider] Questions with performance metrics doing
large inserts into DB
Hi Edward
I'm also coming from a delphi background (using FIB) and we have ported/are
already porting for 5 years our applications to .net. I have noticed indeed
that Delphi/fib is faster then .net provider. But never in the magnitude of
500%. It looked acceptable slower.
Your "write" code seems to be correct and pretty optimal. Usually people
recreate a command each time. I've also done some profiling in the past and
I've noticed that keeping a reference to the parameter doesn't help much. A
slightly slower method, but way less code is to "clear" the parameters and
recreate them. Once again, It is slightly slower, neglectible, but in your
scenario 170 lines of code less.
using (command = new command){
var par = command.Parameters
while (! Eof){
par.Clear();
par.Add(Id);
par.Add(Value);
...
}
}
On the other hand Firebird is an open source database and also the .net
provider. Jiri (the guy supporting this) is getting a few bucks per month to
support this. We, as a company, sometimes sponsor these things by testing,
benchmarking or lazy picking up the bill. Since you have a testing environment
up and running, you might walk the extra mile and help everybody by profiling a
bit deeper? This would benefit you, me and everybody.
Two more things...
1. Your transaction parameters please. How do you create them, what settings?
2. Did you check your "source-read" logic. Cound you maybe fake data, so we
know if it is the "insert" that is slow. Maybe the materializing of the source
record is slow or fetching it?
Looking forward for tackling this thing.
thanks
a
________________________________
From: emendez...@nc.rr.com<mailto:emendez...@nc.rr.com>
To:
firebird-net-provider@lists.sourceforge.net<mailto:firebird-net-provider@lists.sourceforge.net>
Date: Thu, 7 Aug 2014 18:14:43 -0400
Subject: Re: [Firebird-net-provider] Questions with performance metrics doing
large inserts into DB
Hello All,
I have to develop an application that will move "old/stale" data from certain
tables to another FB DB. We already have an existing application that did
something similar to this, but his application is written using Delphi 5 and we
are a .NET shop and wanted to develop newer applications using .NET
technologies so we can reuse our developer resources. Little by Little we have
been migrating off from Delphi5 to .NET.
In our shop we are running various instances of FB on 2.14 Classic on CentOS
5.6. Our Database is larger than 250GB.
In past .NET projects I have used Dapper dot Net and thought that this might
fit the requirements.
We developed a working prototype of what we wanted I had our testers run the
application to see what they thought. To my dismay, they informed me that the
performance was terrible.
In some cases we need to archive millions of rows to the other Database. And it
seemed using dapper was not giving us acceptable results. The users said that
using the old Delphi applications was quicker when archiving data ( that Delphi
application has functionality that also archives different data to other
Databases).
I wanted to have a baseline test so we can compare apples to apples.
I trimmed down the prototype to eliminate Dapper dot net and use straight
ADO.NET for the Inserts using the latest Firebird.NET provider on .NET 4.5.1.
In the application once the data was retrieved I read it one
row at a time, because trying to read in over a million rows into memory would
cause Out of Memory issues. A sample of the code I am using is shown below. As
I looped over the results, I reported on every 1000 rows and calculated elapsed
times every 5000 rows. Running this application from various computers in our
infrastructure to various target DBs the best performance I got was 19 seconds
for 5000 rows.
I then put on my Delphi Hat On and created a small app that did similar to what
the .NET app is doing and the Delphi app’s performance blew away the .NET
performance with a consistent 5000 rows @ 4 seconds. This is Delphi5! >From the
year 1999. I also have to add that I am using the FIBPlus data components from
devrace.
Below is an edited version of my .NET code.
query = @"Select *** Here is my Select Query from the Source DB *** ";
// This is using dapper.net to retrieve the rows I need.
var en = db_source.Query<TABLE_A_DTO>(query, new { ARCHIVE_SET_ID =
m_archiveSetId, ARCHIVE_DATA_TYPE = _tableName }, srcTransaction, buffered:
false);
int totalUpdated = 0;
JobStartTime = DateTime.Now;
try
{
using (FbConnection db_target =
((FbConnection)GetConnection(targetDB)))
{
FbTransaction transaction = db_target.BeginTransaction();
m_isInTransaction = true;
using (FbCommand command = new FbCommand(_updateSQL,
db_target, transaction))
{
command.Parameters.Add("TABLE_A_ID", FbDbType.Integer);
// There are 86 fields in the result set so i do
this 85 more times
command.Prepare();
StepStartTime = DateTime.Now;
_logger.Write("ARCHIVING of " + _tableName + "
Started", 2);
// I get the Enumerator of the results so I can iterate
over them
_iEnumerator = en.GetEnumerator();
while (_iEnumerator.MoveNext())
{
rowsRead++;
var rec = ((TABLE_A_DTO)_iEnumerator.Current);
command.Parameters[0].Value = rec.TABLE_A_ID;
// I do this 85 more times for each
parameter
rowsAffected = command.ExecuteNonQuery();
totalUpdated = rowsAffected + totalUpdated;
if (rowsRead % 1000 == 0)
Console.Write(totalUpdated.ToString() +
"\r");
if (rowsRead % m_recordBufferCount == 0)
{
_logger.Write(string.Format("Archived...
Rows Archived = {0}, Elasped Time {1}", totalUpdated.ToString("N0"),
Utils.GetElapsedTime(DateTime.Now - StepStartTime)), 3);
StepStartTime = DateTime.Now;
}
}
Is the .NET Provider that slow? I am sure that the Database Configuration has
something to do with the performance, but If that is true I expected that the
Delphi Application show performance results on par with the .NET performance.
Am I doing something blatantly wrong code? I would hate to go back to my Boss
and tell them that if performance is a factor we need to continue with Delphi.
If this open source .NET Provider is not the “fastest on the Block”, are there
any other 3rd party Libraries that I can use with .NET that you would
recommend? Even if it a commercial product?
Thanks,
Edward Mendez
------------------------------------------------------------------------------
Want fast and easy access to all the code in your enterprise? Index and search
up to 200,000 lines of code with a free copy of Black Duck Code Sight - the
same software that powers the world's largest code search on Ohloh, the Black
Duck Open Hub! Try it now. http://p.sf.net/sfu/bds
_______________________________________________ Firebird-net-provider mailing
list Firebird-net-provider@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/firebird-net-provider
------------------------------------------------------------------------------
_______________________________________________
Firebird-net-provider mailing list
Firebird-net-provider@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/firebird-net-provider
