Thecus N4200 review

[CopenHagenDK]

I can't believe that the reviewer on the Thecus N4200, Tim Higgins , actually compares NAS's without using reference HDD's between the NAS'es he compares. It really isn't very helpful. For example, the highly efficient Thecus N4200 gets the following comment in the "Pros" and "Cons" paragraph:

Cons: Performance lags other D510 Atom NASes

The real reason for the performance lag is due to the three Barracuda 7200.10 80 GB (ST380815AS) which this particular NAS is shipped with. These drives were developed in back in 2006!!

The review doesn't really tell anything about the true potential of the NAS itself, it merely tells you that it isn't good together with some poor drives from 2006. I'm sure that the Cisco and Qnap currently on top of the leaderboard would behave just as poorly with the same drives.

Looks like someone is trying to make the Thecus N4200 look bad. I would advise the reviewer to use reference drives when testing! And please, don't excuse yourself by saying: "but I only tested what I received" ... most NAS'es comes without drives anyway.

The SmartNetBuilder review is about the only review I've seen that doesn't give the Thecus N4200 the credit it deserves, and the reviewer doesn't even bother mentioning that the poor performance is due to the HDD's this particular sample was shipped with.

 

[thiggins]

If there is someone trying to sabotage performance, it's Thecus, who provided the drives, with the N4200.

I request that all manufacturers provide diskless NASes with their choice of drives.

Since some manufacturers make only diskful NASes, it's neither practical nor useful to use the same drives in BYOD NASes.

CPU performance, then OS tunings are the key factors affecting performance.

 

[teknojnky]

I am not sure I agree with the assertion that it is not practical to use the same drives (or same type) with all your testing.

You could always have a dozen (or however many) standardized 500g drives on hand and use them instead of any supplied drives.

This would provide a level of consistency, as drive performance can affect overall nas performance, even if it may be minimal, and it leaves one less variable to worry about.

Heck, if you have the budget, I would like to see testing with both standardized regular drives, and a set of SSD drives to further isolate performance to the hardware/software design instead of how fast/slow a particular set of disk drives are.

something like these which are cheap and fast (size doesn't matter, as these would be for performance testing).

 

[Osamede]

I am not sure I agree with the assertion that it is not practical to use the same drives (or same type) with all your testing.

You could always have a dozen (or however many) standardized 500g drives on hand and use them instead of any supplied drives.

This would provide a level of consistency, as drive performance can affect overall nas performance, even if it may be minimal, and it leaves one less variable to worry about..

I have to agree on that. SMB tries to claim a position of being more scientific and accurate in testing than the other websites. In that light, it looks pretty bad when the reviewer displays this sort of attitude when such inconsistency is pointed out. Diskless NAS should be compared with one standard set of disks. You never hear of hi fi websites for example saying that they tested a set of speakers with any old wires and music source that the manufacturer sent along in the box.

 

[kamina]

I have to agree. Actually I'll admit I've only read the graphs in some of the reviews, and taken for granted they where set up with the same disks. I guess that explains some performance differences between the reviews.

 

[thiggins]

I have done a quick poll of some of the NAS manufacturers and they say that drives can have a "minor" effect on performance in some NASes. The major determinant of performance is, as I said, processor power and OS tunings (write caching and oplocks primarily).

Not many are going to use SSDs in NASes due to price / performance. There are plenty of cheaper, higher capacity SATA drives (Spinpoint F3, for example) that will allow any NAS that can saturate a Gigabit Ethernet connection to do so.

At any rate, I don't plan any changes in how BYOD NASes are tested.