Consider you have a message of 100 bytes. If it is split in between 2
buffers, you can write it as follows:
[Chunk 1]
65 - 65 bytes to follow
false - incomplete flag
data - 65 bytes
[Chunk 2]
35 - 35 bytes to follow
true - complete flag, this is the end
data - 35 bytes
15 окт. 2016 г. 1:06
Vova,
You still can't write the first buffer to channel before the whole message
is serialized. Are you going to have multiple write buffers? I'm confused :)
-Val
On Fri, Oct 14, 2016 at 1:49 PM, Vladimir Ozerov
wrote:
> Val,
>
> No need to copy on write. You simply
Vova,
I meant the copy on write. To know the length you need to fully marshal
message first. This means that you will always copy the array before
writing to channel. Unless I'm missing something, this eliminates the
purpose of direct serialization.
-Val
On Thu, Oct 13, 2016 at 11:09 PM,
I already tried smth like that when tried to make marshalling parallel.
This is how I tried to reapproach it.
1. allocate set of direct buffers per each session (connection). that was 1
large buffer, but sliced with smaller ones over large offheap array.
2. each thread acquires chunk by chunk
Valya,
Yes, in this design we will copy data into separate buffer on read. But
what is important - it will happen only for message which is split between
buffers.
On Fri, Oct 14, 2016 at 2:33 AM, Valentin Kulichenko <
valentin.kuliche...@gmail.com> wrote:
> Vladimir,
>
> We don't write length
Writes can be optimized even further:
1) Write to *ByteBuffer *as long as there is a place in it.
2) When it is full - invoke a callback which will submit it to the socket,
reset position to 0, and continue marshaling.
This way we can probably get rid of write "state" at all.
On Thu, Oct 13,
Folks,
I went through our so-called "direct serialization" and appears to be not
very efficient to me. We never write message length. As a result we have to
constantly track what was written and what was not, and whether we have a
room for the next write. The same goes for reader. As a result