NASLite Network Attached Storage

Netgear card is using Cat5e cable, 10', 1000Mb light is lit. Using Netgear GS108 gigabit switch but no jumbo frame support. status reports Netgear card is connected in 1000 mode but MTU is set to 1500 (would that change if I had a gigabit switch with jumbo frame support?)

Ran Diskwrigger, got this:

Writing 2000 x 699840 byte frames into file Y:\/framestre
401 frames in 5.00 secs ==> 80.20 FPS (53.53 MB/s)
285 frames in 5.00 secs ==> 57.00 FPS (38.04 MB/s)
341 frames in 5.00 secs ==> 68.20 FPS (45.52 MB/s)
293 frames in 5.08 secs ==> 57.70 FPS (38.51 MB/s)
[...]

Reading 2000 x 699840 byte frames from file Y:\/framestre
59 frames in 5.13 secs ==> 11.51 FPS (7.68 MB/s)
60 frames in 5.22 secs ==> 11.50 FPS (7.67 MB/s)
66 frames in 5.13 secs ==> 12.88 FPS (8.60 MB/s)
[...]


Um - what the hell? I can write at 40MB/s, but only read at 8?

Running NasLite 2.0, box is an Athlon 1600+, 1GB ram, single SATA 7200 RPM drive connected to Adaptec 1210 RAID card, but shows up as JBOD (just a bunch of disks), so it's NOT raided. Could that be the issue, the raid card? I doubt it, right?

I would have expected MUCH better performance than THIS. Yuck!

The 1210 card says RAID but in reality the RAID is based in the drivers and not in an embeded hardware engine on the card. RAID0 and RAID1 are easy to do in software since there is no parity to be generated. It is basicaly data sreering between drives.

As to the poor read performance, you guess is as good as mine. I run a Netgear GA621 card and have no problems at all. I don't know if the card is based on the same chipset though. Ralph uses Netgear hardware to do his testing so it must be good stuff. Have you tried replacing the patch cord yet?

Mike

I know it's not a real RAID card. But when it's installed as JBOD, it should be as fast as any other SATA controller, I assume. THus, I'm assuming it's the network card.

Why is the MTU set to 1500, can anybody tell me how to set it higher, or does it auto-negotiate? One of the things that bugs me is that I don't know how to "break into the shell" on NasLite in order to set files. Do I do it from telnet on a client machine? does that work? How do I edit files on NasLite's box assuming I can telnet in? vi? emacs? some other editing proggie?

NASLite is locked down, there is no shell access that I know of.

Mike

Well SOMEBODY tell me how to set the MTU for the netgear card. I can't believe there's no shell access. weird.

There is no way unless your card has a Flash to store settings in the hardware of the card. You would have to access this using a utility from the makers site. THERE IS NO WAY TO DO THIS IN NL DIRECTLY!

Mike

how do you know that?

I told him.

Well zipwax,

Let me start with this: NASLite-2 is not a general purpose OS and as such is not designed with facilities to allow one to “tinker” with it. It is built as an appliance OS and configures itself automatically to accommodate the available hardware environment. Configuration options are minimal for obvious reasons, making it accessible even to an absolute novice.

That said, considering the context of NASlite, your expectations are probably outside the scope of what one may consider reasonable. If you spend some time and read the various posts on this board, you’ll probably notice that most people that try NASLite also commit to using it. That’s usually because they are happy with performance in both speed and stability. You can also spend some time on Google, where you’d see the same general consensus about what NASLite can do.

Considering the above, it’s easy to see that the problem you are experiencing is likely associated with your particular hardware and/or the way NASLite-2 handles that hardware. NASLite-2 is a 3.5MB operating system that does what it does and supports what it supports with the facilities contained in that tiny footprint. It should not take a genius to deduce that there are hardware configurations that would yield better performance with NASLite than others. On the average however, the stats are clear. NASLite is stable, fast and easy to use, and that’s exactly why people choose to use it.

Anyway, this is partly in response to your email in addition to the above posts. The benchmark tests are not rigged but done on stock DELL machines. We figured that it made since to test a server on hardware designed as such. The results are factual and are pretty good for a “just a toy” as you describe NASLite-2.