Jump to content

ClusterKnoppix Image

Sebastian Smith

Recommended Posts

Howdy folks. Its been a while, I decided to rejoin the Hak5 game.

Quick question; does anyone happen to have a ClusterKnoppix ISO still? Its so old I can not find a mirror.

I have a bunch of old x86 servers taking up space and I decided they need to take up electricity too.

Thanks in advance. Hope to see you at shmoocon!

Link to comment
Share on other sites

ClusterKnoppix will almost certainly not do anything useful for you. Development of the upstream project was abandoned because there was no longer any pragmatic benefit to be had on the commodity hardware of the time.

Your comment surprised me, so I went and searched for a more elaborate explanation for why ClusterKnoppix ceased.

As it turned out way back in 2003, OpenMosix, which underpinned it, wasn't planning on advancing from Linux kernel 2.4 to 2.6 due to the invasiveness of their kernel work and the amount of effort it would take to port that to 2.6 which I can only assume to be the reason why ClusterKnoppix stopped being developed - nobody wanted to miss out on the rather impressive performance improvements provided by Linux 2.6 and they wouldn't be able to deliver it. I couldn't find an actual announcement to that effect though.

OpenMosix itself was abandoned in 2007 (again, Linux 2.6.0 came out in 2003 and they still only had an unofficial patch) mainly because the original project lead came to the conclusion that his project wasn't doing what it set out to do. Their method of clustering was to move tasks to less busy nodes to properly load all nodes in the cluster. This works well when your task is a process. When your task is a thread, much less so, not in the least because multiple threads of a program tend to interact. Clustering a website would suck performance-wise. Clustering a database was next to impossible. Basically, their cluster would only allow you to cluster heavily parallel, relatively long-running and very independent tasks like clustered compiles, bitcoin mining or seti@home/distributed.net type loads. The people that were dealing with these problems were solving them either using MPI and/or just adding more resources to a single machine because the latency involved in moving a task off a system, along with the administration that follows, would tank the performance to the point of unusability (Moshe Bar himself admits the impact was a 100x to 1000x drop in performance).

I wasn't into clusters as much then as I am now, so this little peek down memory lane delivered a lot of insight.

Thanks for that! :smile:

Edited by Cooper
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

  • Recently Browsing   0 members

    • No registered users viewing this page.
  • Create New...