News for hackers. Stuff that matters.

c0t0d0s0

Re:No, ZFS really doesn't need an fsck tool on Saturday March 30, 2013 @04:00AM (#5464810)
When i leverage the copyonwrite-ness (which is more redirect-on-write) of the filesystem to recover from a defunct state of the on-disk state to a known state, a filesystem check is just a suboptimal solution. That what i wanted to express with the article. Of course - most filesystems are not COW and can't use it, and thus the notion that a filesystem check is needed prevails. But at the end a filesystem check is just about forcing a filesystem into a state of metadata correctness, without caring much about the data. I wouldn't count that as a way out, when there are better ways.

I think the situation is pretty much similar to the "shoot the messenger" problem of ZFS. Some people are annoyed that ZFS reports errors because of corruption and blocks access to the data (of course without having any redundancy). However the alternative would be reading incorrect data. What's worse. Knowing that you have to recover data or processing incorrect data without knowing it.

Re:No, ZFS really doesn't need an fsck tool on Saturday March 30, 2013 @02:34AM (#5464670)
The basic problem of filesystem repair is the point that you repair the metadata, but it can do nothing about your data. So when your filesystem enables you to fall back to a known consistent state of metadata and data by the COW, you should fall back to a known consistent state. And not to something that is repaired by a generic tool. How do you know in such a situation that somewhere in you thousands and thousands of files the repair got something wrong. The copy on write behaviour of ZFS has it's advantages.

And as i already wrote in a different comment: If there is a bug in the stuff writing the on-disk state, the bug should be addressed on the exact knowledge of the bug in the code reading the on-disk-state and thus doesn't make assumption what could be halfway correct, but by some piece of code that does the correct with the incorrect on-disk-state.

Re:No, ZFS really doesn't need an fsck tool on Friday March 29, 2013 @02:41PM (#5462272)
Scrubbing is more like reading the block, calculate the checksum. It's the same as stored for the block (not at the block, important difference)? Nice. Not the same checksum. Go to the mirror and fetch the correct one, respectively use an algorithm to calculate the missing block from the raidz1/2/3 set and write it to the disk with the error.
Re:No, ZFS really doesn't need an fsck tool on Friday March 29, 2013 @12:19PM (#5461434)
1. When there is a bug in the code that writes the ZFS stuff, why should the bug be addressed by the fsck code. This would assume, that you know of the bug beforehand, but then you could better fix the bug in the code that writes.

2. When there is a bug in the on-disk-state it should be addressed by the code that reads the data , not by a fsck tool.

2.1. The correction of the bug in the on-disk-state should be done on the basis of the exact knowledge about the bug and not by a generic check tool.

3. Repair is always based on assumptions. Those could be correct or incorrect. The more you know about the problem that led to the repair-worthy state, the more probable the assumptions are correct.

4. What is the reasoning behind the argument "when your metadata is corrupt , that the data is correct" and so you could repair metadata corruption without problems. It sounds more sensible to fall back to the last known correct and consistent state of metadata and data, based on the on-disk-state represented by the pointer structure of the ueberblock with the highest transaction group commit number with a correct checksum . The Transaction Group rollback at mount does exactly this.

Re:Thoughts of a sysadmin starting to get old ... on Saturday July 25, 2009 @02:55AM (#723067)
When everybody thinks like that, only bureaucrats define standards. Lack of standards let to many problem: I´m just thinking about the amount of power supplies i have to carry. A power supply for the company mobile, a power supply for the iPhone (or at least the USB cable), a power supply for the notebook. I have three small switches from three vendors needing three separate voltages (a standard would be nice here, for example for a central power supply feeding all this stuff). My USB hard drives have their own power supply. The 1.5 TB Maxtor from the special offer have a different one than the 2 1TB Fujitsu-Siemens drives. Standards are bureaucratic, but they help ... but we have to work on them, to enable them to serve us. And at the end it´s the same at power supplies as with Unix standards.
Re:Thoughts of a sysadmin starting to get old ... on Friday July 24, 2009 @05:51PM (#722510)
So instead to help to make them faster, just ignore them?
Re:Thoughts of a sysadmin starting to get old ... on Friday July 24, 2009 @05:50PM (#722505)
Perhaps because portability wasn't just an afterthought in your development.
Re:Thoughts of a sysadmin starting to get old ... on Friday July 24, 2009 @05:27PM (#722462)
At least in Solaris the kernel interfaces doesn't change from minor to minor release rendering existent binary drivers useless ;)

But that aside: At the end the text is not about Linux, about Solaris or AIX, FreeBSD, NetBSD, OpenBSD. It's about portability. Either we have to accept that there is a lock-in to Linux for many opensource development and stop all this portability stuff.

The article is about saying: Hey, use gmake if you want ... but keep into consideration, that the target system may be something else than Debian Lenny.

The next thing: I would like to see a standard way to install software on Linux. 17 distinct packages for Linux, just one of Mac OS X and 2 for (Open)Solaris. A standard would be nice here ... really.

And you have to take the following into consideration: Outside of the enthusiasts group something like binary compatibility guarantees something important for the admins. Improvements are good, features are good. But not at the price of breaking every bridge behind. At the end this hurts the Unix franchise, as the Microsoft franchise is simply a much larger market. We can't afford to fragmentize the Unix business by being to GNU/Linux centric. Deal with it. Solaris won't disappear, the *BSD won't disappear. But we should be able to ensure, that all can participate from open source development. And one factor is further development of the Unix standards to keep them current with the technological development, and not just cooking it's own soup.

Re:Thoughts of a sysadmin starting to get old ... on Friday July 24, 2009 @05:06PM (#722422)
No ... i don't want a pony ;) I just want to be able to to compile code on Solaris, Linux, one of the BSD without having to make vast adaptions (or hope that someone else made it). It isn't even about libraries. Of course, you need libraries. It's about compliance to standards, and that Linux shouldn't invent it's own standard and instead work with the established standard bodies. Linux isn't the centre of the universe.

Obviously you didn't understand the point of the article. It's not about hating Linux or GNU ... it's just about the ability to use open source software outside the Linux Community without ending to throw out all this Linuxisms and GNUisms

At the end you can see it as the plea, that portability isn't just an afterthought ... and given the state of some makefiles and hard coded things in some open source developments you get the impression that it's a not-thought-about it at all.

PS: I'm from the Opensolaris side of the unix universe.

Re:Thoughts of a sysadmin starting to get old ... on Friday July 24, 2009 @02:19PM (#722073)
1. The problem is a different one: You develop for a 1-2 core machine and afterwards this stuff is destined to work on real servers. I can´t count how often i found massive scaling problems in systems you wasn´t able to find on a notebook, but was pretty obvious on larger system. Sometimes i was able to remove the problem, but often we replicated the development plattform by using 1-2 socket servers.

2. Obviously GNU make has advantages. I don´t have anything against GNU make. I just hate this Linux centric view. But from my perspective it would be more feasible to extend the standards (for example SUSv5 ;) ) to keep programs more portable between all Unixes and derivates. Free Software should be operating system agnostic.

3. I have to correct you: HPUX and OSF/1 is mainly on life life-support because of HP ill-fated decisions with PA-RISC and Alpha in favour of Itanic. AIX isn´t a toy for testing ... in the real world there are many systems running those operating sytems like Solaris and AIX. You just have too look at the market data: Sun and IBM still sell systems worth billions of $.

4. Yes gmake is in Solaris ... because of the dependence of many build procceses. But i´m still thinking it would be better to define unixwide standards that include GNU make. At the moment the behaviour is similar to Microsoft. We have our LSB .. that´s a standard ... the bad thing ... it´s a extended version of POSIX, SUS and so on. I think people should ask, why there is the need for a extended own standard instead of extending the existing standards.