C# sqlite/NoSQL advise with a choice

Hi all,

could you tell me which implementation you should choose.
I need to save quite a lot of data.
But all must be saved in separate files.

is it worth to try sqlite and db4o.
I would like something in common between sqlite and mongoDB

Thanks in advance
October 8th 19 at 01:21
4 answers
October 8th 19 at 01:23
why do you need storage in separate files? why files?
October 8th 19 at 01:25
the fact that data should then copy to another media
most often in manual. Separate files are needed for different groups, and months.
And if you make a field a normal base, and the transfer of the produce with the help of import export Livo by means of standard DBMS or to write what is their mechanisms. - Sydney.Bartoletti commented on October 8th 19 at 01:28
October 8th 19 at 01:27
sqlite is one of the slowest implementations on record (on the average win the car from 100ms per transaction), even ms access faster, but reading smart, plus compatible high (if you want a separate file, then you need to transfer from machine to machine? and there's sqlite for ALL platforms)

nosql in General difficult to compare with sql, the highest speed (in the order above sql), but the key -> value creates limitations, and makes sense mostly for document-oriented databases, i.e. if you have 1->m it will have to serialize the list and to monitor the integrity of ourselves.
what can you say about Sql Server Compact Edition
what is the speed of read/write? - Sydney.Bartoletti commented on October 8th 19 at 01:30
stackoverflow.com/questions/5249240/sql-server-ce-4-0-performance-comparison - modesto_Beahan commented on October 8th 19 at 01:33
SQLITE is fast and write and read. Just by default, each team creates and commits its transaction. The transition to group transaction and is pending a commit timeout to save the case. - jaunita commented on October 8th 19 at 01:36
Dear opponent, I understand you all like sqlite, but I quite clearly wrote it about the time of execution of the transaction and not the individual teams.

I can think of a thousand and one use sqlite when a write operation can be performed without worrying about whether or not they performed, a little less — when I call time, during which I enough that the data structure was not consistent (delayed commit), but in the General case is more advantageous for me to change backend on something faster, the benefit of options dark, though the same mysql embended.

p.s. for tasks, the records are rare or do not exist — I would first recommend sqlite, read implemented quite quickly (was the project base has grown to 14GB, slow sequential write and a bunch of minor search query — only positive impressions) - Norbert.Konopelski commented on October 8th 19 at 01:39
October 8th 19 at 01:29
If You require further fetch operation and data analysis?
If not, then can directly to a file should write.
this is the thing that you want
in addition, the data will be a lot
another drawback of the binary serialization of objects is that it can easily "go wrong", it is then or to dig hexeditor, or score data that will be very bad. - Sydney.Bartoletti commented on October 8th 19 at 01:32

Find more questions by tags SQLSQLiteNoSQLMongoDBC#