• A Great friend to the HardForum with a great kid that he is trying to get a scholorship to continue his schooling. Please give hime a vote! Only 24 hours left! Thanks.
    If you have an VOTE FOR KEENAN!

Backblaze POD v2

I'd love to put something like that together and use ZFS. Seems they're using ext4
 
Very nice, these have improved drastically over their old models.

btw, Ryan711, check out some of the comments below the article to see why they chose Ext4 over ZFS, if you haven't already.

Lots of good info and great for data centers.
 
I wonder what their rebuild time is. Port multipliers are überslow.
 
Backblaze's design just does not make sense unless you are building a bunch of them. For making one or two large file servers, it is better to spend a little more for better performance and convenience.
 
Still no Ecc. Not good. Imgine how many errors happen there every day.
 
Still no Ecc. Not good. Imgine how many errors happen there every day.

I would actually guess not as much as you would think, from the memory. Remember they are only using 8GB of RAM.

I'm also thinking of getting one of the cases and trying to find SAS backplanes that work.
 
That article has made me very glad to have picked Hitachi as the drive for my build.

EDIT: A good guess as to why they don't want to go with OpenSolaris (as per the comments in the link) is probably due to the fact that it is no longer being actively developed and OI probably just doesn't have the same amount of people backing it as Linux does.
 
Last edited:
Still no Ecc. Not good. Imgine how many errors happen there every day.

I would say 0 per decade although I have not obviously tested that to confirm. At work I have run memtest86+ on systems with over 8 GB of memory for over a week without a single bit error. Actually if I saw a single error I would have RMA'd the ram.
 
Actually they do see a lot of errors

(Disclaimer: I work at Backblaze) - If you really care about data, you *MUST* have end-to-end application level data integrity checks (it isn't just the hard drives that lose data!).

Let's make this perfectly clear: Backblaze checksums EVERYTHING on an end-to-end basis (mostly we use SHA-1). This is so important I cannot stress this highly enough, each and every file and portion of file we store has our own checksum on the end, and we use this all over the place. For example, we pass over the data every week or so reading it, recalculating the checksums, and if a single bit has been thrown we heal it up either from our own copies of the data or ask the client to re-transmit that file or part of that file.

At the large amount of data we store, our checksums catch errors at EVERY level - RAM, hard drive, network transmission, everywhere. My guess is that consumers just do not notice when a single bit in one of their JPEG photos has been flipped -> one pixel gets every so slightly more red or something. Only one photo changes out of their collection of thousands. But at our crazy numbers of files stored we see it (and fix it) daily.
 
The Crucial memory they list is unbuffered ECC. But it has always been unclear to me whether the Intel i3-540 even supports ECC memory. On ark.intel.com, it does not have a YES or a NO for ECC support. As compared to the Xeon X3430 which does explicitly support ECC:

http://ark.intel.com/compare/46473,42927

As far as I can tell, they could be operating ECC memory without any ECC actually in use.
 
That article has made me very glad to have picked Hitachi as the drive for my build.

EDIT: A good guess as to why they don't want to go with OpenSolaris (as per the comments in the link) is probably due to the fact that it is no longer being actively developed and OI probably just doesn't have the same amount of people backing it as Linux does.

was thinking the same thing.
have 2 NAS' with hitachi drives (one with 1TB 7K1000.c HD's & one with 2TB 5K3000 HD's).

based on their own internal audits, just show how rubbish the Seagate & WD drives are.
just hope WD doesn't mess with the hitachi storage division too much.
 
Back
Top