You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@etch.apache.org by "scott comer (JIRA)" <ji...@apache.org> on 2008/11/20 18:10:44 UTC
[jira] Created: (ETCH-20) need a more flexible buffering scheme to
handle different transports
need a more flexible buffering scheme to handle different transports
--------------------------------------------------------------------
Key: ETCH-20
URL: https://issues.apache.org/jira/browse/ETCH-20
Project: Etch
Issue Type: Improvement
Components: csharp-binding, java-binding
Affects Versions: 1.0.1, 1.0.0
Reporter: scott comer
Assignee: scott comer
Fix For: 1.0.2
FlexBuffer has a pretty wide interface and is a concrete class. underneath the hood it has a single byte array which grows to meet demand, but a single large byte array per connection times thousands of connections doesn't scale. also thrashes the heap.
solutions:
1) make FlexBuf an interface.
2) breakout data processing methods into separate input / output interfaces (much like java DataInput and DataOutput).
3) make a concrete impl of FlexBuf which works for TcpConnection and big endian formats.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (ETCH-20) need a more flexible buffering scheme to
handle different transports
Posted by "scott comer (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/ETCH-20?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
scott comer updated ETCH-20:
----------------------------
Fix Version/s: (was: 1.0.2)
> need a more flexible buffering scheme to handle different transports
> --------------------------------------------------------------------
>
> Key: ETCH-20
> URL: https://issues.apache.org/jira/browse/ETCH-20
> Project: Etch
> Issue Type: Improvement
> Components: csharp-binding, java-binding
> Affects Versions: 1.0.0, 1.0.1
> Reporter: scott comer
> Assignee: scott comer
>
> FlexBuffer has a pretty wide interface and is a concrete class. underneath the hood it has a single byte array which grows to meet demand, but a single large byte array per connection times thousands of connections doesn't scale. also thrashes the heap.
> solutions:
> 1) make FlexBuf an interface.
> 2) breakout data processing methods into separate input / output interfaces (much like java DataInput and DataOutput).
> 3) make a concrete impl of FlexBuf which works for TcpConnection and big endian formats.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (ETCH-20) need a more flexible buffering scheme to
handle different transports
Posted by "scott comer (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/ETCH-20?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
scott comer updated ETCH-20:
----------------------------
Assignee: (was: scott comer)
> need a more flexible buffering scheme to handle different transports
> --------------------------------------------------------------------
>
> Key: ETCH-20
> URL: https://issues.apache.org/jira/browse/ETCH-20
> Project: Etch
> Issue Type: Improvement
> Components: csharp-binding, java-binding
> Affects Versions: 1.0.0, 1.0.1
> Reporter: scott comer
>
> FlexBuffer has a pretty wide interface and is a concrete class. underneath the hood it has a single byte array which grows to meet demand, but a single large byte array per connection times thousands of connections doesn't scale. also thrashes the heap.
> solutions:
> 1) make FlexBuf an interface.
> 2) breakout data processing methods into separate input / output interfaces (much like java DataInput and DataOutput).
> 3) make a concrete impl of FlexBuf which works for TcpConnection and big endian formats.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.