eCluster(tm) Server Release 2.6 - from www.xgforce.com


	PRODUCT DESCRIPTION:

	eCluster(tm) is a highly scalable Intelligent Load Balance
	Cluster, which can be scaled up to 1024 Internet cluster
	groups and each contains 1024 cluster nodes(1024X1024).
	With round trip time load balance and fail safe Algorithm,
	one can clusterize any OSes on any CPUs, such as NT, Novell, UNIXes,
	etc, with infinite standbys. The support of load balance or fail
	safe mode, ensured non-stoppable and fast business transactions
	for SQL Database Servers, such as Oracle, MS SQL, Informix,
	MySQL, Postgress, etc.

	Load balance algorithms include: CPU Load, Weighted Load, 
	VM Usage, Round Trip Time, CPU Usage, Traffic Allocation, etc. 

	The eCluster(tm) Server provides Network Traffic Distribution, 
	Network Failsafe, CFS(tm), and Large Network Management features
	for both Internet and Intranet, where one could distribute 
	network traffic among large number of networked heterogeneous 
	servers and workstations. It's able to detect busy servers and 
	and crashed network services. Once detected, eCluster(tm) will
	route traffic to server/workstation that is non-busy and with the
	most available resources. It came with an email subsystem that 
	all resource shortages and malfuctions will be reported 
	automatically in real time to administrators via email capable 
	pagers/cell phones. It's able to clusterize any servers and/or 
	workstations on the Internet by virtually grouping them together. 

	Network Traffic Distribution speeds up Internet services. Web 
	Servers and FTP Servers will have distributed traffic. Network 
	traffic will be routed only to the server/workstation that is up, 
	with non-busy and non-crashed corresponding service(port), and 
	with most available resources. For example, if the Server Nodes
	is short in virutual memory, file system space, or too many
	network connection, the traffic will be routed to the one with
	least used Server Nodes. Should malfunction occurs, an email 
	alert will be sent to a prespecified email address where 
	administrators can be notified. The malfunctional Server Node
	will be taken offline. It provides better performance and faster
	response to users with fail safe assurance.

	The eCluster(tm) failsafe capability allows non-interruptible 
	Internet services and Intranet services. Administrators may 
	take down systems temporarily for server and/or Internet service 
	maintenance without interrupting Internet or Intranet users. 
	It is a true High Availability System providing large scale of 
	redundancy.
	
	Large Network Management helps administrators to view and
	monitor the whole network at their finger tip(a single image). 
	Once the network is eClustered, statistics of each server and 
	workstation can be seen from a single console. Alert messages 
	will inform administrators resource shortages, such as memory 
	shortage, hard drive space shortage, etc. Administrators 
	will be alerted periodically if the shortage condition or 
	failure event is not fixed. The server or workstation that 
	is short in resource will be taken off-line automatically by 
	the Cluster Master Daemon(Clmd) and connection request 
	will be directed to the server/workstation with most 
	available resource and least latency. Administrators can be 
	notified by pager if you have email capable pagers, or
	cell phones.

	eCluster(tm) also provides file system clustering across all 
	cluster nodes in any given virtual cluster group. This Clustered 
	File System Daemon (CFSD) provides file system virtual replication 
	and synchronization in "real time" through high speed encrypted and
	compressed data communication channels. The CFSD(tm) is distributed 
	along with the eCluster(tm) distribution. It's able to cluster nodes 
	dispersed on the Internet and is OS and computer architecture independent.

	Please read man pages after installation for more detailed 
	features. For any questions, send mail to info@xgforce.com.


	eClUSTER(tm) DESIGN DECISIONS AND IMPLEMENTATION NOTES:

	The eCluster(tm) is a fully distributed cluster with unrestricted level
	of distributed redundancy. It offers several advantages over the device
	implementation. Device implementation came with two flavors, the DNS
	version and the NAT based version.

	The device implementation(DNS or NAT based) is limited in its hardware.
	It's not as flexible as the software implementation, which may run on
	any OSes, any machine architectures. In comparison, the software 
	approach is an OPEN implementation. One may choose his own OS, his own 
	hardware configuration, even CPU architectures, in a low cost fashion, 
	even assemble his own hardware. A device is limited to its own enbeded 
	OS, the limited power of the enbeded CPU, and the amount of memory 
	shipped with the device.

	The NAT based device implementations has its drawback where it has to
	be setup as a router where all traffic must going through this device
	and so the load balance and fail safe will be possible. The NAT device
	will have to spend extra amount of CPU power to:

		1. Inspect each packet going in and out to the Internet. 
		2. Search through its connection tables - If there're
		   1 million connections to various local network servers,
		   the search will take a long time.
		3. Modify each packet in/out of the Internet.

	This way of setup made the device became a single point of bottleneck.
	Extra packet processing slowdowns the bandwith of the vital Internet 
	Link of the whole local network. The correct setup is to free the 
	router from any extra packet proessing work as much as possible. 
	The NAT device tend to be a bottleneck of the Internet connection 
	given the same hardware configuration.


	TECHNICAL SUPPORT:

	Life time free support is provided. Free configuration and free
	training is also provided to both eval users and registered users. 
	If you have installation or configuration problems, send email to 
	support@xgforce.com.


	PORT NOTES:

	The SUN SPARC port works for Solaris 2.6, 2.7, and 2.8. The SUN Intel
	port will be soon abailable.

	The FreeBSD port is compiled on 4.1, and 2.2.2 stable. The 
	4.1 port doesn't work on FreeBSD 4.0 or any previous, but maybe works
	on later versions such as the 5.x. However, they're not tested (only if 
	kernel structures is not changed). The 2.2.2 port should be working on 
	all 2.x.
	
	The LINUX port works on Redhat 6.0, 6.1, 6.2. These include
	kernel 2.2.5-15 and kernel 2.2.12-20. For most recent LINUX kernels, 
	it might be working as well even though not tested. The port should
	work with other LINUX distributions as long as it is the same kernel. 

	In case you are using a new version of the OS, simply let us know.
	We might port over to it and send the software over to you.


	INSTALL:

	Please read the updated version of the configuration instruction 
	file at: ftp.xgforce.com/pub/eCluster/2X4/INSTALL.TXT.


	MISC:

	Visit our Web site at www.xgforce.com for new version releases.
	Or visit our ftp site at ftp.xgforce.com. We are working hard to 
	further improve our software. Your advise, comments, are very 
	valuable. Please feel free to contact us at info@xgforce.com.   

	All *.TXT files located in this directory contains up to date info.
	It's constantly updated. The *.TXT files in the *.tgz distribution
	may be outdated. 

	It's recommended to download the lastest version of this software. 
	CFS(tm) is only distributed in the eCluster(tm) 2.4 release or later.