Slashdot Mirror

OpenSolaris is a good choice I believe. It can be downloaded for free from http://opensolaris.org/os/ or you can request a free copy to be shipped to your door. Use CDE for desktop environment and you're laptop will be rekindled!

Read this PDF. In particular, page 13:
Measurements at CERN
Wrote a simple application to write/verify 1GB file
* Write 1MB, sleep 1 second, etc. until 1GB has been written
* Read 1MB, verify, sleep 1 second, etc.
Ran on 3000 rack servers with HW RAID card
After 3 weeks, found 152 instances of silent data corruption

Reading/Writing 1MB/sec for three weeks has 152 SILENT errors. They don't mention detected errors, but it has to be more than the silent ones.

Wow. I love your FUD. If you're going to lie, at least make it seem truthful.

Lacking in file system utilities (yes, fsck IS necessary even on healthy filesystems, especially on desktops and portables)

Why no fsck? And if you really feel the need to do something:

zpool scrub <pool_name>

License-incompatible with anything worth running it on, other than Solaris itself... which is NOT worth running (see #1 above)

What you mean to say is "Some Operating Systems whose merits can be debated are license incompatible with the license of ZFS." FreeBSD can implement ZFS. Why can't Linux? Because of its license, not that of ZFS.

See here and Sun's silly response as to why they neglected to provide fsck support for ZFS in their OS.

From a previous mini-thread on the matter. See the full thread for plenty of other examples in defense of fsck utilities on "robust" filesystems.

"ZFS has checksums and will find errors, but only will be able to self-heal the errors in a redundant configuration. On a single disk, ZFS will find the error thanks to checksums but will not be able to recover your data. Since ZFS was mainly designed for systems that will use redundant configurations, it may have sense there, but desktops are not never going to do such things. IMO the ZFS people were a bit elitist here - "let's going to build a filesystem so good that we won't need a fsck". But in the real world you _are_ going to need a fsck util. Only in excepcional and very rare cases, but you're going to need it."

We need another layer, between the block layer and the filesystem layer [...] which is simply concerned with allocating some amount of space [...] Filesystems could sit above this layer

This is exactly how ZFS is designed:

• SPA - Storage Pool Allocator: allocates blocks from all the devices in a storage pool
• DMU - Data Management Unit: consumes blocks from the SPA and exports objects (flat files)
• ZPL - ZFS POSIX Layer: makes DMU objects look like a POSIX file system

See page 7 of this ZFS presentation for a comparison with the traditional block device model.

One of the differences I can find between btrfs and ZFS is that ZFS explicitely avoided a fsck utility, and btrfs is explicitely designed with features designed to make fsck even more powerful than it's on usual filesystems like ext3. In btrfs, data structures have "back references", and the fsck can be used while the filesystem is mounted.

IMO, this is a a btrfs advantage. ZFS has checksums and will find errors, but only will be able to self-heal the errors in a redundant configuration. On a single disk, ZFS will find the error thanks to checksums but will not be able to recover your data. Since ZFS was mainly designed for systems that will use redundant configurations, it may have sense there, but desktops are not never going to do such things. IMO the ZFS people were a bit elitist here - "let's going to build a filesystem so good that we won't need a fsck". But in the real world you _are_ going to need a fsck util. Only in excepcional and very rare cases, but you're going to need it.

Of course that doesn't makes ZFS a bad filesystem, but it's an advantage for btrfs and linux.

http://www.opensolaris.org/os/project/pkg/ is claimed to be as good as apt-get.

I do not think it is, at least yet, as good and robust. After all apt-get has been around for years.

I'd like to have Bluetooth, and overall better support for peripherals in OpenSolaris. This will, obviously, take a long time, if ever. After that I'd ditch Linux for good.

The future of Solaris on the desktop is not as exciting as that of Ubuntu

I think it is as exciting. But then, I am excited by new (free) operating systems ...

The latest OpenSolaris has "pgk" which performs about same as apt-get http://www.opensolaris.org/os/project/pkg/.

My next home machine will be OpenSolaris (NFS/Samba server with raidz2).

I agree with those people advocating ZFS.
It is insanely simple (and powerful) to administer.
Shrink, grow, create, snapshot, delete, manage filesystems intuitively with power and flexibility seldom seen in ANYTHING else.
Seriously, read the PDF below to see what you're giving up by using anything else:
http://opensolaris.org/os/community/zfs/docs/zfs_last.pdf
http://opensolaris.org/os/downloads/
http://www.genunix.org/

There are several OS distributions besides OpenSolaris / Solaris that use ZFS, NEXENTA, FreeBSD, Schillix, Belenix, et. al.;
NEXENTA uses most of the open source type of desktop / utility applications instead of the Solaris style ones, I think others do that to some degree as well. IDK if MythTV and your HW drivers run under FreeBSD 7 but if it did that'd be perfect.

There is a simple solution to getting it to be the storage drive for LINUX, though -- run a virtual machine under LINUX and in the VM run FreeBSD or OpenSolaris or whatever with a ZFS filesystem given full control of the devices making up the ZFS storage pool. Run the LINUX OS off either a distinct drive or partition so that it doesn't get interfered with by the ZFS / other OS.
Share the ZFS over NFS between the OS versions in the host/guest via their VMed networking and that should have plenty of performance for HTPC use. You'll lose about 1GB RAM or so for doing the VM thing, and you'd probably need 1 core of CPU free to handle it and still give the other OS good performance, but those aren't entirely unreasonable tradeoffs today.

I agree with those people advocating ZFS.
It is insanely simple (and powerful) to administer.
Shrink, grow, create, snapshot, delete, manage filesystems intuitively with power and flexibility seldom seen in ANYTHING else.
Seriously, read the PDF below to see what you're giving up by using anything else:
http://opensolaris.org/os/community/zfs/docs/zfs_last.pdf
http://opensolaris.org/os/downloads/
http://www.genunix.org/

There are several OS distributions besides OpenSolaris / Solaris that use ZFS, NEXENTA, FreeBSD, Schillix, Belenix, et. al.;
NEXENTA uses most of the open source type of desktop / utility applications instead of the Solaris style ones, I think others do that to some degree as well. IDK if MythTV and your HW drivers run under FreeBSD 7 but if it did that'd be perfect.

There is a simple solution to getting it to be the storage drive for LINUX, though -- run a virtual machine under LINUX and in the VM run FreeBSD or OpenSolaris or whatever with a ZFS filesystem given full control of the devices making up the ZFS storage pool. Run the LINUX OS off either a distinct drive or partition so that it doesn't get interfered with by the ZFS / other OS.
Share the ZFS over NFS between the OS versions in the host/guest via their VMed networking and that should have plenty of performance for HTPC use. You'll lose about 1GB RAM or so for doing the VM thing, and you'd probably need 1 core of CPU free to handle it and still give the other OS good performance, but those aren't entirely unreasonable tradeoffs today.

I admire your optimism. But if you watch threads like this one and others that go back three years, I think you'll be disappointed. Sun's original story was "no one should ever need to shrink a pool". In the face of a *huge* number of responses to the effect that "yes, but in the real world you do", their response for a couple of years was "well then you just don't understand ZFS". Then when that didn't work they switched to "we've been working on it for years, we just want to get it right so be patient for another few years and in the mean time just keep buying more disks".

The folks at Sun are quite smart, but they're not infallible. My guess is that some early design decisions about ZFS made shrinking extremely hard, and they're having a hard time living that down given how much they crowed about ZFS being "the last word in filesystems". Eighteen months ago I was super-excited about ZFS since it just had one more feature to go before it fit my needs. Now I don't expect to ever see that feature, at least not before some better alternative comes along with equivalent features and a more Linux-friendly license (btrfs+lvm?).

But note that shrinking pools is not yet supported. It's almost the #1 wishlist item though, so it shouldn't be too long coming.

While ZFS can be RAM hungry, the 8GB suggested by another poster is probably exaggerated, depending on your workload. A light ZFS workload runs fine on a 2GB box (which is what I run it on). 64-bit architecture is recommended for best performance.

More info at the OpenSolaris ZFS Community.

http://www.opensolaris.org/os/community/zfs/

there are regional OSUGs around the place supported by Sun , freebies and lectures, look them up and see if you're near one

I have had driver problems with Linux as well. Does it have such a site?

PS. You can also check the forums at http://opensolaris.org/os/

1. Error reporting paths are not reliable (in fact it's notoriously terrible). Drive failure prediction is not reliable.

2. The data path is not reliable. That includes media, drive controller, firmware, buffers, cabling, controller, host RAM, and other host subsystems. Cosmic rays. Whatever might flip a bit anywhere in your system.

3. Given the foregoing, ZFS detects ALL errors between media and application, and (if redundancy is available) corrects them.

Traditional RAID doesn't even know which side of a mirror is the good side, if they don't match. Some RAID systems do checksumming but this doesn't protect you against errors in the rest of the datapath. And isolated storage subsystems like NetApp, no matter how sophisticated, also do not protect the whole longer more vulnerable datapath.

ZFS is unique in doing that, by design. Join the mailing list or study the material at opensolaris.org. It may change the way you think about storage.

I have been a ZFS user for a while and know a lot of its internals. Let me comment on what you said.

checksums really only help in detecting errors.

Not in ZFS. When the checksum reveals silent data corruption, ZFS attempts to self-heal itself by rewriting the sector with a known good copy. Self-healing is possible if you are using mirroring, raidz (single parity), raidz2 (dual parity), or even a single disk (provided the copies=2 filesystem attribute is set). The self-healing algorithm in the raidz and raidz2 cases is actually interesting as it is based on combinatorial reconstruction: ZFS makes a series of guesses as to which drive(s) returned bad data, it reconstructs the data block from the other drives, and then validates whether this guess was correct or not by verifying the checksum.

checksums have limits on how many errors they can detect.

All the ZFS checksumming algorithms (fletcher2, fletcher4, SHA-256) generate 256-bit checksums. The default is fletcher2 and offers very good error detection (even errors affecting more than 256 bits of data) assuming unintentional data corruption (the fletcher family are not a cryptographic hash algorithms, it is actually possible to intentionally find collisions). SHA-256 is collision-resistant therefore it will in practice detect all data corruptions. It would be computationally infeasible to come up with a corrupted data block that still matches the SHA-256 checksum.

A good intro to the ZFS capabilities are these slides

With Raid5 .. two drives fail and your done.

Which is why some people do RAID 6 instead. Or you could keep doing RAID 5, but periodically scrub your array to detect failing disks ASAP, like ZFS scrubbing. What, your RAID implementation doesn't implement scrubbing ? Maybe you should blame it instead of blaming RAID 5, and consider switching to ZFS :-)

No, here is why GPL was not chosen:

If you wanted a copyleft license, why didn't you just use the GPL or LGPL?

We needed an open source license that allowed files released under the license to be linked with files released under other licenses. While a license like LGPL would allow this for dynamically-linked code, we also needed to be able to release software that statically links source files available under different licenses. In addition, we wanted to allow others to add externsions to OpenSolaris with different license terms. This was only possible under a license like the MPL; however, we could not use the MPL because it is not a "template" license allowing reuse by others. Consequently, we crafted a variant of the MPL, taking the opportunity to make it a template license as a step towards reducing license proliferation for others finding themselves in a similar position.

The other reason for the creation of the CDDL has to do with software patents:

What does the CDDL say about patents?

The CDDL provides an explicit patent license for code released under the license. This means that you can use, modify, and redistribute code released under CDDL without worrying about any patents that the contributors of the code (including Sun) might have on the contributed technology. The license also includes a provision to discourage patent litigation against developers by revoking the rights to the code for anyone initiating a patent claim against a developer regarding code they have contributed.

The reasons that the GPL is incompatible with the CDDL are very complicated and nuanced but in large part have to do with the patent clauses. It is clear why a company such as Sun needs such clauses. So if it were not for the stupidity of the existence of software patents a CDDL-like license could have been created that would have been compatible with the GPL minus some other niggles. Because of the existence of software patents and the need for such clauses the CDDL is incompatible with the GPL. It is too bad about the GPL being so restrictive about adding clauses that protect the copyright holders.

Really the spirit should have been that if CDDL source is used in another project that this project needs to be open, but then all sorts of real world complications get in the way. That is basically the spirit of the GPL as well. Sun was against anything BSD-like where another company could take their source and create a closed source product.

As noted below, you propose using SSD as a buffer. Well, this is done today with ZFS (available in Solaris, FreeBSD, Mac OS X). It's called L2ARC. Look here for details

Open Solaris - http://opensolaris.org/

FreeBSD - http://freebsd.org/ (excellent bridged IPless firewall capability)

NetBSD - http://netbsd.org/

Yeah, that must be why TrustedBSD is copying SELinux (just like opensolaris)...

People claims SELinux is difficult, but they often don't understand how insanely powerful it is....

You should read this thread, it has more information http://in.opensolaris.org/jive/thread.jspa?messageID=146731&#146731

It's not Debian. Debian has had the ability to fully encrypt the root partition during installation since Sarge I think. Etch for sure. Ubuntu can do it too with the alternate installer. OpenSuse and Slackware have excellent docs on how to get / file encryption. Disk Encryption is essential for laptops and removable media in 2008. If Solaris wants to get adopted by government and financial sectors for use on laptops it will need to have some form of serious disk encryption. To be fair to the OpenSolaris people there are two teams working on encryption solutions but I think they lag well behind Linux or even Windows (Truecrypt) solutions. Two in development projects: Crypto in the lofi(7D) driver (a bit like dm-crypt on Linux or FileVault on MacOS X): http://opensolaris.org/os/project/loficc/ due to integrate soon. and ZFS Crypto which is still in development but due to integrate this summer. http://opensolaris.org/os/project/zfs-crypto/ However neither of these provide for an encrypted root filesystem as they aren't full disk encryption solutions. However with ZFS Crypto all of your home directory and other datasets (filesystems) with sensitive data can be encrypted. I for one welcome my Sun Microsystems overlords...actually I am glad to see another alternative to Windows becoming more accessible to the masses. I have my copy in bittorrent now ready to install in my [Sun Microsystems] Virtualbox 1.6.0 Congratulations to the Project Indiana Team!

It's not Debian. Debian has had the ability to fully encrypt the root partition during installation since Sarge I think. Etch for sure. Ubuntu can do it too with the alternate installer. OpenSuse and Slackware have excellent docs on how to get / file encryption. Disk Encryption is essential for laptops and removable media in 2008. If Solaris wants to get adopted by government and financial sectors for use on laptops it will need to have some form of serious disk encryption. To be fair to the OpenSolaris people there are two teams working on encryption solutions but I think they lag well behind Linux or even Windows (Truecrypt) solutions. Two in development projects: Crypto in the lofi(7D) driver (a bit like dm-crypt on Linux or FileVault on MacOS X): http://opensolaris.org/os/project/loficc/ due to integrate soon. and ZFS Crypto which is still in development but due to integrate this summer. http://opensolaris.org/os/project/zfs-crypto/ However neither of these provide for an encrypted root filesystem as they aren't full disk encryption solutions. However with ZFS Crypto all of your home directory and other datasets (filesystems) with sensitive data can be encrypted. I for one welcome my Sun Microsystems overlords...actually I am glad to see another alternative to Windows becoming more accessible to the masses. I have my copy in bittorrent now ready to install in my [Sun Microsystems] Virtualbox 1.6.0 Congratulations to the Project Indiana Team!

I stand by my original statements 100% (I'm a certified SAP basis engineer on Sun equipment).

The source code (which remaps linux systems calls to open solaris and fudges inconsistencies)

Info on installing debian (it's designed for RedHat based linux, so it's slightly painful ... though possibly out of date).

Brand Z info

Overview of linux support

I haven't tried it, but there shouldn't be much overhead/performance loss.

The source code (which remaps linux systems calls to open solaris and fudges inconsistencies)

Info on installing debian (it's designed for RedHat based linux, so it's slightly painful ... though possibly out of date).

Brand Z info

Overview of linux support

I haven't tried it, but there shouldn't be much overhead/performance loss.

The source code (which remaps linux systems calls to open solaris and fudges inconsistencies)

Info on installing debian (it's designed for RedHat based linux, so it's slightly painful ... though possibly out of date).

Brand Z info

Overview of linux support

I haven't tried it, but there shouldn't be much overhead/performance loss.

I'm still using Solaris 10 for a project I'm working on but am looking to move it to OpenSolaris before release.

Project Crossbow is one of the projects I wish was currently available now. It looks like the easiest way to set up virtual switches and networks which is a great feature to use along with zones. Right now I'm using a hack I found online to do this. Crossbow is a lot easier and integrated with SMF. I haven't really had time to really focus on making a management script for the hack yet. It's not too hard but I have been focusing on other areas.

Sun is all about Java so many of the tools like IPS are written in it.

I can tell you haven't even tried solaris 10, but give it a swig. Before solaris 10 I wrote (often rightly) wrote of Sun. Why would I pay a premium for something FreeBSD can do for free and outperforms it? The hardware is cool (see coolthreads processors...it's hyperthreading done right), it's affordable, and it's innovative. It may not be compelling enough to switch from linux or whatever if all you use from a desktop is firefox and thunderbird, but there is actually some VERY cool stuff in there. Don't write it off. There's a reason FreeBSD is taking in a lot of these features.

Anyone here know what's so special about the Image Packaging System? I found the homepage, but it didn't really explain how it differed from traditional packaging methods. (More annoyingly, it didn't even explain that intriguing name!) A quick check of Wikipedia doesn't offer much help, either. Anyone know the scoop on this (new?) system?

I can understand if people are not comfortable with using ReiserFS in light of what has happened. I never used it, so I cannot offer any opinions on whether it was any good

If people are looking for something different, why not ZFS?

Sure, I know that Sun's commitment to Open Source is inconsistent and potentially suspect, but ZFS looks to be an impressive piece of technology.

So, why not ZFS?

Sun didn't just make vague statements to me about OpenSolaris; they made promises about it being an open development project. That's the only way they could get someone like me to provide free labor for their benefit. Given Sun's recent track record on breaking promises, another one doesn't surprise me at all.

This well is poisoned; the company has consumed its own future and any pretense that the projects will ever govern themselves (as opposed to being governed by whatever pointy-haired boss is hiding behind the scenes) is now a joke. Sun should move on, dissolve the charter that it currently ignores, and adopt the governing style of MySQL. That company doesn't pretend to let their community participate in decisions, and yet they still manage to satisfy most of their users. Let everyone else go back to writing code/documentation for hire.

It seems Sun is doing everything in its power to alienate a developer community.

-Wouldn't let the opensolaris board call the project opensolaris. Probably a legal quagmire of their own creation. The consequences of that lead to this resignation. http://mail.opensolaris.org/pipermail/ogb-discuss/2008-February/004488.html

-There's this gem, most of which I don't pretend to understand. The punchline is on the bottom. http://cryptnet.net/mirrors/texts/kissedagirl.html

-There's this gem, where even Ian Murdock links in suggesting the difficulty is happening above his level. http://ianskerrett.wordpress.com/2008/02/22/a-solution-for-suns-os-community-problems/#comment-17418

Off the top of my head:

* Quicksilver?
* BitTorrent?
* ZFS?

A whole different problem which I encountered myself several times is that everyone was allowed to spout of several amounts of FUD without ever being told that they did so. Worse yet; the stories sometimes never changed, but it /did/ have a nice "opensolaris" URL attached which has confused people numerous of times. A classic example right: here.

I've seen several people rant about how it wasn't possible to tell (Open)Solaris not to automatically update the hostname and as such you had to change a whole lot of system scripts in order to get this behaviour. This specific (officially accepted so it seems) howto even described on several pages how to best approach the patching of the system scripts.

Total Madness! the only thing people had to do was to change /etc/default/dhcpagent. I've written the author, I've responded in an OpenSolaris forum that this approach is utter nonsense and I've even seen this URL pop up in a few newsgroups here and there.

And this is just one obvious example... For me it was enough to stop taking the whole thing seriously (note: I am a veteral Solaris admin & user). And something tells me I wasn't the only one...

Where have you been? Java source code has been available for a long time but after years of people complaining that it wasn't "free enough", Sun fully released Java under a GPL 2 years ago.

O.K. so it's LGPL So what, so is Gtk, most of GNOME and probably 80% of what you and joe-sixpack considers to be "opensource" in "Linux". GPL is just one license. GPL was never fully tested in court and doesn't provide patent indemnity as CDDL does. I'd be happier if Java, OpenOffice and MySQL were CDDL but there would be too much gnashing of teeth from the Linux creationists.

OpenSolaris is licensed under CDDL. Look here for an FAQ which explains in simple terms why CDDL is superior to GPL.

OpenSolaris, from which came features such as Dtrace and ZFS ,

OpenOffice.org,

GlassFish,

OpenSSO, or

OpenDS

(and probably several others that I missed) aren't really open source?

Thanks for enlightening me. After scratching all these projects off my list, it looks like you're right. Sun hardly open sources anything!

Where have you been? Java source code has been available for a long time but after years of people complaining that it wasn't "free enough", Sun fully released Java under a GPL 2 years ago.

O.K. so it's LGPL So what, so is Gtk, most of GNOME and probably 80% of what you and joe-sixpack considers to be "opensource" in "Linux". GPL is just one license. GPL was never fully tested in court and doesn't provide patent indemnity as CDDL does. I'd be happier if Java, OpenOffice and MySQL were CDDL but there would be too much gnashing of teeth from the Linux creationists.

OpenSolaris is licensed under CDDL. Look here for an FAQ which explains in simple terms why CDDL is superior to GPL.

and it has been around for years: http://www.opensolaris.org/os/community/security/projects/tx/

Last I checked, you couldn't just add a device to ZFS. Instead you have to create a whole new vdev to add to the pool. So if I had a ZFS on 2 500GB drives, one data one parity, and I wanted to add another 500 gb drive so I'd have 2 data and 1 parity, I can't do that. I'd have to buy two whole 500gb drives and have 2 data and 2 parity drives. And even then, those 2 parity drives are kind of wasted being split among vdevs like that, since it would be possible for me to lose both drives in one vdev and lose data. Whereas if I started with 2 parity drives for the whole thing, I could lose 2 drives with no data loss.

Of the Operating Environments you mentioned, source is available. The last "hold-out" was Microsoft -- even they make source licenses available now.

HP: see http://licensing.hp.com/slm/swl/view.slm?page=source (VMS, Tru64)

Solaris: completely open-source, see http://opensolaris.org/os/

IBM: not sure about them -- older releases of IBMs mainframe OS came with source, so I expect that z/OS comes with source. I *haven't* personally seen the source for AIX.

In general, OSs have ALWAYS come with source; back in the early '80s, for example, Digital VMS came with source (by default on microfiche AFAIR). The "closed source" OS was debuted by CP/M, and carried forward by MSDOS.

xVM is Sun's name for their solaris-integrated Xen hypervisor.

Xsun though is only an Xserver, with no device specific knowledge - the
drivers for various graphics devices for Xsun come from 3 other groups
at Sun - SPARC Graphics, x86 Platform Drivers, and Sun Ray.

nbd is nice for some stuff but lacks fault-tolerance. Of course, you can run RAID, possibly several levels (say, a raid-6 on top of raid-1 or something) on top of nbd devices to trade space for fault-tolerance as much as you want, but you still lack flexibility. The advantage to RAID-over-nbd, on the other hand, is of course that you can do that right now if you want :] (And yes, the nbd server shouldn't be overly hard to run on Windows, one would think; it's rather simple...)

A better solution would work on a bit higher level, though. If a host goes down, it would be desirable to flexibly duplicate its data (from other mirrors and/or parity data) onto others. Possibly such a system could be created on top of nbd as well. Hell, maybe ZFS with an NBD pool could someday hack that, but seems to me they'd need to work out at least bug 4852783 first.

That is what OpenSolaris WebStack is doing. Every thing preconfigured and out of the box. and yes Sun does make a commercial JSP container shared by Application server and Web Server disc:IWFS