SI
SI
discoversearch

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor.  We ask that you disable ad blocking while on Silicon Investor in the best interests of our community.  If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.
Technology Stocks : EMC How high can it go? -- Ignore unavailable to you. Want to Upgrade?


To: pirate_200 who wrote (12156)2/13/2001 4:49:58 PM
From: Gus  Read Replies (3) | Respond to of 17183
 
You benchmark with the exact same configuration to highlight how much better your software is, that is why. You don't understand that...

Why? NTAP's configuration is a well-known single point of failure design that can only support one RAID level. It also carries the risk of putting too many disk drives (168) behind one processor. A few years ago, IBM agreed to resell StorageTek's Iceberg technology because its venerable RAMAC controller technology couldn't keep up with the rapid increases in storage capacity. Notice how EMC puts up to 100 disk drives behind 4 processors in the IP4700? Consider that the disk drives will be going from 73GB to 181GB to 500GB and 1 Terabyte in the next few years.

No wonder you go around in circles a lot. You think all storage systems are like commodity PCs and servers. You need to distinguish between core (datacenter) and edge (branch/department/workgroup) storage systems. Then you need to distinguish between low-end, mid-range (5-100 servers) and high-end systems using the type of enterprise applications that need to be supported as a guide.

Symmetrix is a high-end core system. Celerra is a high-end edge system. Clariion FC4700 (SAN) is a mid-range core system while the Clariion IP4700 (NAS) is a mid-range edge system. NTAP's filers have traditionally been mid-range edge systems. You only end up looking confused when you compare mid-range edge systems with high-end core systems without even understanding the type of applications supported by each system.

Not surprisingly, you fixate on a one-dimensional speed benchmark that everybody, including EMC, uses as a marketing tool more than anything else. By the way, the different RAID levels indicate different type of protection schemes for different applications and do not indicate that one is better than the other. RAID 0 is the only level that does not offer any redundancy and is often used for multi-media applications while RAID 1 is the only level that offers 100% redundancy and maximum protection. Except for the IP4700 (RAID 5), all of EMC's boxes offer multiple RAID levels and support for multiple data formats.

NAS hangs off a segment of a highly probabilistic network that is frequently congested and typically operates at less than 30% of wirespeed with up to 50% processor loads so the REAL-WORLD performance of any NAS depends heavily on the availability of the network. Benchmarks are for the gullible.

Currently, Fast Ethernet (100 Mbps) is the mainstream corporate network; although, many companies are already starting to move to Gigabit Ethernet (1 Gbps) which may become the mainstream corporate network technology in 3-5 years. After Gigabit Ethernet comes 10 Gigabit Ethernet (10 Gbps).

10 Gigabit Ethernet, which is expected to be standardized in early 2002, requires a major infrastructure upgrade because it requires optical cabling and is expected to be at least 10x the cost the today's Gigabit Ethernet. Gigabit Ethernet is currently 6x the cost of Fast Ethernet. Watching those price differentials over time gives you a more realistic idea of how fast those corporate networks will become and obviously, how fast those NAS systems will become.

Why do you think all the NAS vendors are backpedaling furiously into SANs? LOL.