Development timeline

2006-07-03T17:09:23Z

Sameer:

==SRM Interface==

* '''7th August''' - deployable as tarball with upload, download, stat, ls, delete, mkdir

* '''15th September''' - fully tested

==Chord==

* '''August 21st''' - Persistent storage used for Chord nodes

==L-Store server==

* '''24th July''' - delete, overwriting working <Srm dependency> '''Priority 2'''

* '''24th July''' - A version of lserver and client that use chord instead of relational database to manage name space. '''Priority 1'''

* Expiration of exnodes

* Authorization and Authentication:

User specifies a username and a password that the lstore server uses to determine whether they are allowed to access a file. User can specify rights on their files using the lstcp client (chmod). There is a list of users in the system.

* Clyde testing framework

* Checksums

* Arbitrary storage device

==Thor==

* '''July 24th''' - Client can simulate a

thor-client thor://server.edu:9000/filename localfilename
thor-client localfile thor://server.edu:9000/newfilename

* '''July 24th''' - client can send data to buffer and buffer gets written to tape if there on time or buffer size events.

* '''August 28th''' - releaseable <alpha> tarball that can be used to provide functionality above.

Main Page

2006-06-08T01:16:56Z

Sameer:

<big>'''MediaWiki has been successfully installed.'''</big>

Consult the [http://meta.wikipedia.org/wiki/MediaWiki_User%27s_Guide User's Guide] for information on using the wiki software.

== Getting started ==

* [http://www.mediawiki.org/wiki/Help:Configuration_settings Configuration settings list]
* [http://www.mediawiki.org/wiki/Help:FAQ MediaWiki FAQ]
* [http://mail.wikipedia.org/mailman/listinfo/mediawiki-announce MediaWiki release mailing list]

== LStore ==

* LStore pages

* [[LStore client installation]]

This is under construction

* [[LStore FAQ]]

== REDDNET Testing ==

* [[Network Characterization Tests]]

* [[L-Store Upload Structured Testing]]

* [[Sockets versus Channels]]

Network Characterization Tests

2006-06-08T01:15:12Z

Sameer:

Bandwidth testing: All tests are an average of five runs

File Size used: 260 MBytes

=Local to Local Tests=

umich-umich (MBytes/second) (Depot was configured to receive max 64 threads.
The figure values increase to the 90's once we reconfigure depot with 120 threads)

{| class="wikitable" style="text-align:center"
* Umich-Umich (MBytes/second)
|-
!bgcolor=red | 1 || bgcolor=blue | Thread || bgcolor=grey | 70.652
|-
!bgcolor=red | 10 || bgcolor=blue | Threads || bgcolor=grey | 82.114
|-
!bgcolor=red | 20 || bgcolor=blue | Threads || bgcolor=grey | 82.582
|-
!bgcolor=red | 40 || bgcolor=blue | Threads || bgcolor=grey | 85.216
|-
!bgcolor=red | 60 || bgcolor=blue | Threads || bgcolor=grey | 81.658
|-
!bgcolor=red | 100 || bgcolor=blue | Threads || bgcolor=grey | 75.644
|-
|}

----------------------------------------
{| class="wikitable" style="text-align:center"
* caltech-caltech (MBytes/second)
|-
!bgcolor=red | 1 || bgcolor=blue | Thread || bgcolor=grey | 81.321
|-
!bgcolor=red | 10 || bgcolor=blue | Threads || bgcolor=grey | 100.08
|-
!bgcolor=red | 20 || bgcolor=blue | Threads || bgcolor=grey | 99.690
|-
!bgcolor=red | 40 || bgcolor=blue | Threads || bgcolor=grey | 99.160
|-
!bgcolor=red | 60 || bgcolor=blue | Threads || bgcolor=grey | 98.820
|-
!bgcolor=red | 100 || bgcolor=blue | Threads || bgcolor=grey | 97.260
|}

----------------------------------------
{| class="wikitable" style="text-align:center"
* Fermi-Fermi (MBytes/second)
|-
!bgcolor=red | 1 || bgcolor=blue | Threads || bgcolor=grey | 74.200
|-
!bgcolor=red | 10 || bgcolor=blue | Threads || bgcolor=grey | 95.870
|-
!bgcolor=red | 20 || bgcolor=blue | Threads || bgcolor=grey | 95.553
|-
!bgcolor=red | 40 || bgcolor=blue | Threads || bgcolor=grey | 93.290
|-
!bgcolor=red | 60 || bgcolor=blue | Threads || bgcolor=grey | 91.872
|-
!bgcolor=red | 100 || bgcolor=blue | Threads || bgcolor=grey | 90.940
|-
|}

=Over the Network Tests=

{| class="wikitable" style="text-align:center"
*UMich to Fermi
|-
!bgcolor=red | 1 || bgcolor=blue | Thread || bgcolor=grey | 8.580 MB/s
|-
!bgcolor=red | 10 || bgcolor=blue | Threads || bgcolor=grey | 66.290 MB/s
|-
!bgcolor=red | 20 || bgcolor=blue | Threads || bgcolor=grey | 68.820 MB/s
|-
!bgcolor=red | 40 || bgcolor=blue | Threads || bgcolor=grey | 66.84 MB/s
|-
!bgcolor=red | 60 || bgcolor=blue | Threads || bgcolor=grey | 67.45 MB/s
|-
!bgcolor=red | 100 || bgcolor=blue | Threads || bgcolor=grey | 66.81 MB/s
|}

---------------------------------------

{| class="wikitable" style="text-align:center"
* Umich to Fermi to 4 depots
|-
!bgcolor=red | 10 || bgcolor=blue | Threads || bgcolor=grey | 69.22 MB/s
|-
!bgcolor=red | 20 || bgcolor=blue | Threads || bgcolor=grey | 68.89 MB/s
|-
!bgcolor=red | 40 || bgcolor=blue | Threads || bgcolor=grey | 68.24 MB/s
|-
|}

These numbers probably are limitations on the lors_upload ability to
pump out data: 1 or 4 target depots see same throughput.

---------------------------------------

{| class="wikitable" style="text-align:center"
* Fermi to Caltech
|-
!bgcolor=red | 1 || bgcolor=blue | Thread ||bgcolor=grey | 0.83MB/s
|-
!bgcolor=red | 10 || bgcolor=blue | Threads ||bgcolor=grey | 7.04MB/s
|-
!bgcolor=red | 20 || bgcolor=blue | Threads ||bgcolor=grey | 6.95 MB/s
|-
!bgcolor=red | 40 || bgcolor=blue | Threads ||bgcolor=grey | 7.06 MB/s
|-
|}

---------------------------------------

{| class="wikitable" style="text-align:center"
* UMich to Caltech
|-
!bgcolor=red | 1 || bgcolor=blue | Thread ||bgcolor=grey | .75 MB/s
|-
!bgcolor=red | 10 || bgcolor=blue | Threads ||bgcolor=grey | 6.37 MB/s
|-
!bgcolor=red | 20 || bgcolor=blue | Threads ||bgcolor=grey | 6.35 MB/s
|-
!bgcolor=red | 40 || bgcolor=blue | Threads ||bgcolor=grey | 6.40 MB/s
|-
|}

-----------------------------------------

=Large file transfer...1.7GB (Only doing this on relatively fast sites)=

{| class="wikitable" style="text-align:center"
* Umich to Fermi
|-
!bgcolor=red | 1 || bgcolor=blue | Thread ||bgcolor=grey | 8.74 MB/s
|-
!bgcolor=red | 10 || bgcolor=blue | Threads ||bgcolor=grey | 49.46 MB/s
|-
!bgcolor=red | 20 || bgcolor=blue | Threads ||bgcolor=grey | 49.43 MB/s
|-
!bgcolor=red | 40 || bgcolor=blue | Threads ||bgcolor=grey | 49.4 MB/s
|-
|}

----------------------------------

=Capricorn Test=

{| class="wikitable" style="text-align:center"
*Capricorn-ibp.accre
|-
!bgcolor=red | 1 || bgcolor=blue | Thread ||bgcolor=grey | 24.89 MB/s
|-
!bgcolor=red | 10 || bgcolor=blue | Threads ||bgcolor=grey | 28.63 MB/s
|-
!bgcolor=red | 20 || bgcolor=blue | Threads ||bgcolor=grey | 28.44 MB/s
|-
!bgcolor=red | 40 || bgcolor=blue | Threads ||bgcolor=grey | 26.74 MB/s
|-
!bgcolor=red | 60 || bgcolor=blue | Threads || bgcolor=grey | 27.51 MB/s
|-
!bgcolor=red | 100 || bgcolor=blue | Threads || bgcolor=grey | 27.37 MB/s
|-
|}

------------------------------------------------

=iperf testing between caltech and michigan=

{| class="wikitable" style="text-align:center"
*caltech-depot01.ultralight.org to umich-depot01.ultralight.org
|-
!bgcolor=red | 1 || bgcolor=blue | Thread ||bgcolor=grey | 0.93 MB/s
|-
!bgcolor=red | 10 || bgcolor=blue | Threads ||bgcolor=grey | 10.4 MB/s
|-
!bgcolor=red | 20 || bgcolor=blue | Threads ||bgcolor=grey | 21.0 MB/s
|-
!bgcolor=red | 40 || bgcolor=blue | Threads ||bgcolor=grey | 42.4 MB/s
|-
!bgcolor=red | 60 || bgcolor=blue | Threads || bgcolor=grey | 63.4 MB/s
|-
!bgcolor=red | 100 || bgcolor=blue | Threads || bgcolor=grey | 105.0 MB/s
|-
|}

--------------------------------------------------------------

{| class="wikitable" style="text-align:center"
*umich-depot01.ultralight.org to caltech-depot01.ultralight.org
|-
|-
!bgcolor=red | 1 || bgcolor=blue | Thread ||bgcolor=grey | 0.93, 0.94 MB/s
|-
!bgcolor=red | 10 || bgcolor=blue | Threads ||bgcolor=grey | 10.3, 10.7 MB/s
|-
!bgcolor=red | 20 || bgcolor=blue | Threads ||bgcolor=grey | 20.8, 20.9, 20.6 MB/s
|-
!bgcolor=red | 40 || bgcolor=blue | Threads ||bgcolor=grey | 11.4, 28.7, 7.02, 20.4 * MB/s
|-
!bgcolor=red | 60 || bgcolor=blue | Threads || bgcolor=grey | 10.4, 35.3, 11.6, 43.5 ** MB/s
|-
!bgcolor=red | 100 || bgcolor=blue | Threads || bgcolor=grey | 58.4, 23.8, 14.1, 22.4 ** MB/s
|-
|}

This table shows results of individual trials rather than an average to the show
behavior of this connection.

* = Erratic iperf results, some of the threads die before the test completes,
bandwidth varies greatly from trial to trial.
** = Same as above plus the test fails to complete some of the time.

------------------------------

=Vandy Caltech client (ultralight) IBP depot (Non-Ultralight) Test Results=

{| class="wikitable" style="text-align:center"
*Vandy Ultralight-caltech-depot01
|-
!bgcolor=red | 1 || bgcolor=blue | Thread ||bgcolor=grey | 0.78 MB/s
|-
!bgcolor=red | 10 || bgcolor=blue | Threads ||bgcolor=grey | 6.51 MB/s
|-
!bgcolor=red | 20 || bgcolor=blue | Threads ||bgcolor=grey | 6.61 MB/s
|-
!bgcolor=red | 40 || bgcolor=blue | Threads ||bgcolor=grey | 6.69 MB/s
|-
!bgcolor=red | 60 || bgcolor=blue | Threads ||bgcolor=grey | 6.48 MB/s
|-
!bgcolor=red | 100 || bgcolor=blue | Threads ||bgcolor=grey | 6.62 MB/s
|-
|}

------------------------------------------------

------------------------------

=Vandy Caltech Ultralight Test Results=

{| class="wikitable" style="text-align:center"
*Vandy Ultralight-caltech-depot02
|-
!bgcolor=red | 1 || bgcolor=blue | Thread ||bgcolor=grey | 0.72 MB/s
|-
!bgcolor=red | 10 || bgcolor=blue | Threads ||bgcolor=grey | 6.37 MB/s
|-
!bgcolor=red | 20 || bgcolor=blue | Threads ||bgcolor=grey | 6.34 MB/s
|-
!bgcolor=red | 40 || bgcolor=blue | Threads ||bgcolor=grey | 6.44 MB/s
|-
!bgcolor=red | 60 || bgcolor=blue | Threads ||bgcolor=grey | 6.49 MB/s
|-
!bgcolor=red | 100 || bgcolor=blue | Threads ||bgcolor=grey | 4.40 MB/s
|-
|}

------------------------------------------------

=Vandy Caltech Ultralight Test Results=

{| class="wikitable" style="text-align:center"
*Vandy Non-Ultralight-caltech-depot02 (IBP depot with ultralight kernel on it)
|-
!bgcolor=red | 1 || bgcolor=blue | Thread ||bgcolor=grey | 0.77 MB/s
|-
!bgcolor=red | 10 || bgcolor=blue | Threads ||bgcolor=grey | 6.51 MB/s
|-
!bgcolor=red | 20 || bgcolor=blue | Threads ||bgcolor=grey | 6.62 MB/s
|-
!bgcolor=red | 40 || bgcolor=blue | Threads ||bgcolor=grey | 6.69 MB/s
|-
!bgcolor=red | 60 || bgcolor=blue | Threads ||bgcolor=grey | 6.67 MB/s
|-
!bgcolor=red | 100 || bgcolor=blue | Threads ||bgcolor=grey | 6.51 MB/s
|-
|}

------------------------------------------------

=Vandy Caltech client (non-ultralight) IBP Depot (non-Ultralight) Test Results=

{| class="wikitable" style="text-align:center"
*Vandy Non-Ultralight-caltech-depot01 (non ultralight IBP depot)
|-
!bgcolor=red | 1 || bgcolor=blue | Thread ||bgcolor=grey | 0.78 MB/s
|-
!bgcolor=red | 10 || bgcolor=blue | Threads ||bgcolor=grey | 6.58 MB/s
|-
!bgcolor=red | 20 || bgcolor=blue | Threads ||bgcolor=grey | 6.72 MB/s
|-
!bgcolor=red | 40 || bgcolor=blue | Threads ||bgcolor=grey | 6.69 MB/s
|-
!bgcolor=red | 60 || bgcolor=blue | Threads ||bgcolor=grey | 6.63 MB/s
|-
!bgcolor=red | 100 || bgcolor=blue | Threads ||bgcolor=grey | 6.49 MB/s
|-
|}

------------------------------------------------

Main Page

2006-06-08T01:11:43Z

Sameer: /* LStore */

<big>'''MediaWiki has been successfully installed.'''</big>

Consult the [http://meta.wikipedia.org/wiki/MediaWiki_User%27s_Guide User's Guide] for information on using the wiki software.

== Getting started ==

* [http://www.mediawiki.org/wiki/Help:Configuration_settings Configuration settings list]
* [http://www.mediawiki.org/wiki/Help:FAQ MediaWiki FAQ]
* [http://mail.wikipedia.org/mailman/listinfo/mediawiki-announce MediaWiki release mailing list]

== LStore ==

* LStore pages

* [[LStore client installation]]

* [[Sockets versus Channels]]

This is under construction

* [[LStore FAQ]]

== REDDNET Testing ==

* [[Network Characterization Tests]]

* [[L-Store Upload Structured Testing]]

Sockets versus Channels

2006-06-08T01:09:11Z

Sameer:

This page shall contain information and updates on comparison of network performance when using [http://java.sun.com/j2se/1.4.2/docs/api/java/net/Socket.html Sockets] (java.net.Socket) and when using the new [http://java.sun.com/j2se/1.4.2/docs/api/java/nio/channels/Channel.html Channel] introduced in Java 1.4 ([http://java.sun.com/j2se/1.4.2/docs/api/java/nio/channels/SocketChannel.html java.nio.channels.SocketChannel]).

==External Links==
* Sun's page on the new I/O APIs, [http://java.sun.com/j2se/1.4.2/docs/guide/nio/ here].

ReddNet - User contributions [en]

Development timeline

Main Page

Network Characterization Tests

Main Page

Sockets versus Channels