A first look at NZFS and replacing unRAID with NZFS’s Transparent RAID (tRAID)

Solarnet said:
Brutalizer,

You know, this type of language is completely uncalled for. I sure nobody wants to here this kind of language. If you have a problem with his response please take it to PM.
Click to expand...
We did PM about his behavior, and he apologized. Now he does this again. That is why I am maybe over reacting. You know, when someone does things again and again, then you tend to react. You know nothing of our earlier history, so to you it may seem that I over react.

But sure, I will take it PM next time.
 
TeeJayHoward said:
This. There is a certain degree of respect required in public postings between users on this forum. I'm not a moderator, but I would say that your post has overstepped your bounds. Specifically, I would like to point out this section from the rules:
....Mutual respect and civilized conversation is the required norm.
Click to expand...
Yes, the reason I got upset, is because he broke against this rule you quote. "Mutual respect". Where is his respect, when he does this again and again? No respect. Of course I get pissed! Anyone would. If he had respected my knowledge and my contributions to the debate, I would not have been pissed off again.
 
brutalizer said:
We did PM about his behavior, and he apologized. Now he does this again. That is why I am maybe over reacting. You know, when someone does things again and again, then you tend to react. You know nothing of our earlier history, so to you it may seem that I over react.

But sure, I will take it PM next time.
Click to expand...

You're right I have no knowledge of your past with him, But is it fair to me or anyone else to have to come here and read vulgar language and name calling? It doesn't put you in a good light. You could have worded your response a lot better.
 
Solarnet said:
You're right I have no knowledge of your past with him, But is it fair to me or anyone else to have to come here and read vulgar language and name calling? It doesn't put you in a good light. You could have worded your response a lot better.
Click to expand...
I hear you, and agree with what you say. I am not unreasonable. For my harsh language, I apologize. I would not like children to read things like that. Maybe I should edit away the worst stuff.


Anyway, back to topic: there are lot of research on data corruption that shows that adding checksums does not suffice, to make a safe storage solution. So what makes NZFS safe?



EDIT: Ah, my post got deleted. Well, it is ok. But I am not going to accept disrespect from Odditory a third time. The next time I will let moderators handle him.
 
brutalizer said:
I hear you, and agree with what you say. I am not unreasonable. For my harsh language, I apologize. I would not like children to read things like that. Maybe I should edit away the worst stuff.
Click to expand...

Bravo, Glad to hear it. It takes a man to be a man....:D
 
Next one to take this thread off-topic gets busted, so take that into consideration before you do a self inflicted wound.
 
brutalizer said:
Is it a bad thing, to try to bring in academic research papers into the discussion?
Click to expand...

Of course not.
The degree and e-penis war is entertaining, and I am guilty for enjoying a little bit of drama. :D

brutalizer said:
...
Anyway, back to topic: there are lot of research on data corruption that shows that adding checksums does not suffice, to make a safe storage solution. So what makes NZFS safe?
...
Click to expand...
If you follow the thread, you will find my statement about me saying nothing yet of the the file system aspect of NZFS.
Basically, you are criticizing on empty. ;)

Now, there is theory and then there is practical implementation.
So, please download FlexRAID's RAID-F and prove how it fails to detect datarot under any intended usage.
What's the value of something that only happens in theory and never in practice?
 
spectrumbx said:
If you follow the thread, you will find my statement about me saying nothing yet of the the file system aspect of NZFS.
Basically, you are criticizing on empty. ;)
Click to expand...
Ok, you say nothing on the file system aspect of NZFS. But how can you then claim NZFS detects bit rot? If the underlying filesystem does not detect bit rot (which it doesnt) how can NZFS do it? If the ground is shaky, how can you claim the entire building is safe?


Now, there is theory and then there is practical implementation.
So, please download FlexRAID's RAID-F and prove how it fails to detect datarot under any intended usage.
Click to expand...
So it works fine, only under intended usage? If there are some problems, it might not be safe anymore?
 
brutalizer said:
Ok, you say nothing on the file system aspect of NZFS. But how can you then claim NZFS detects bit rot? If the underlying filesystem does not detect bit rot (which it doesnt) how can NZFS do it? If the ground is shaky, how can you claim the entire building is safe?



So it works fine, only under intended usage? If there are some problems, it might not be safe anymore?
Click to expand...

:confused:

/I give up...
 
Ok. But I really do hope you succeed, and wish you the best of luck. The more competition in the storage area, the better for us the customers. You are going to create something really ground breaking some day! :)
 
It was my understanding from reading the description that NZFS uses ZFS as its underlying filesystem, is this incorrect?
 
leeleatherwood said:
It was my understanding from reading the description that NZFS uses ZFS as its underlying filesystem, is this incorrect?
Click to expand...

Incorrect. It is its own self-contained system. Few pics of interface (Create RAID menu):

TRAID.png


Create-RAID.png
 
I believe brutalizer is asking how the detection of data corruption and how to fix it is implemented. You can't only checksum data blocks, you need to checksum the metadata too, and a sum of the metadata. So everything has to be checksummed all the way up to a superblock. And if something goes wrong, ZFS will rebuild the exact data from the parity silently and return the correct data to the process asking for the specific data.
 
olavgg said:
I believe brutalizer is asking how the detection of data corruption and how to fix it is implemented. You can't only checksum data blocks, you need to checksum the metadata too, and a sum of the metadata. So everything has to be checksummed all the way up to a superblock. And if something goes wrong, ZFS will rebuild the exact data from the parity silently and return the correct data to the process asking for the specific data.
Click to expand...

saying data corruption only, this is a big(mess) topic.
when you put in detail on what "data corruption" that would be good to understand.
 
This all sounds pretty good. The only thing I'm wondering is how far it is. If it's still very much under development it might still be years before it is useable.
 
Why name it "NZFS" unless you're trying to get it confused with ZFS? It doesn't have anything to do with it. If I registered as "nspectrumbx" on the forum, wouldn't that strike you as a bit weird?
 
sukosevato said:
This all sounds pretty good. The only thing I'm wondering is how far it is. If it's still very much under development it might still be years before it is useable.
Click to expand...

The full stack will be out in a few months.
However, the Transparent RAID feature will be out next month.
 
NZFS TRAID is now in UAT (User Acceptance Testing).
It should finalize in the next few weeks.
 
You must log in or register to reply here.
Back
Top