Xsan Release Delayed

[MacRumors]

In an article on CNET News (http://news.com.com/Apples+xSAN+falls+behind+schedule/2100-1012_3-5498360.html) Apple has said it will miss the deadline for shipping its Xsan storage management software. As recently as last month, Apple reiterated the software would ship in the fall of 2004. The delay in the software was previously rumored (http://www.macrumors.com/pages/2004/12/20041208135018.shtml), along with possible upgrades to the Xserve line of Apple servers.

[AoWolf]

hmm

[iostream.h]

I wonder what the problem could have been?

[AliensAreFuzzy]

They really need to finish this up. People are waiting for this

[oldMac]

I would suspect that this is something that's very difficult to write. Combine that with the very new hardware that they're working with and I would guess that some unexpected bugs have crept up.

Worst thing they could do is to release this with problems. You can't mess around with data integrity.

[nagromme]

That was supposed to be my Grandmother's Christmas present :( I already wrote the card and everything :mad:

[savar]

I would suspect that this is something that's very difficult to write. Combine that with the very new hardware that they're working with and I would guess that some unexpected bugs have crept up.

Worst thing they could do is to release this with problems. You can't mess around with data integrity.

This stuff is pretty cutting edge, and to make it rock-solid *and* put a pretty face on it (ie. Apple style) I'm sure requires a ton of code audits.

When they add this to the supercomputing clusters...that will be nuts. The question is: will enterprises take notice of this "enterprise" solution?

[msbsound]

Been in contact with apple constantly on this. Our business rep has no idea when it will be released cause up until 2 weeks ago he was swearing it was days away. He now tells us he hopes the end of January, but not much confidence in his voice. We are on the fence between that and some other fibre channel solutions so I know he is itching to get us the info...

[x86isslow]

Perhaps they plan to release it at the Stevenote, which is just a few weeks off now. They'll likely also release those 2.3 Ghz XServes which they'd previously limited to certain large-volume buyers: (read VT, Colsa). The upcoming XServe release is probably what's holding things up.

[Nucular]

Would it be largely beneficial to Apple to delay the release of the software until they had it bundled with some hardware? That is, I can't see how the development process could be so out-of-touch that they couldn't get a working version out the door. I optimistically suspect that this is partially intentional, based on a delay from some hush-hush project.

[virividox]

hmm

haha do i see a post count comment :)

anyway, everytime apple sets a release date i never hold my breath, its like just ad 1 month or 2 to the release date cuz apple always hits delays and snags!!! and even if it is released, it doesnt mean its shipping on time haha

[Daveway]

Coming from Apple, What doesn't get delayed? :confused:

[swissmann]

They really need to finish this up. People are waiting for this

How many? Doesn't this target a pretty small market? Still the delay isn't welcome.

[MattG]

Apple? Delays? What?

[Wonder Boy]

blame motorola! :p

[Rod Rod]

How many? Doesn't this target a pretty small market? Still the delay isn't welcome.

The market is relatively small but Xsan can still make huge waves. This is a product that will become almost as important as Final Cut Pro, and actually make FCP more important. At the moment one of the biggest advantages of Avid over FCP is the availability of SAN solutions (Avid yes, FCP no).

As for as earlier comments about this delay being related to a hardware release, that's nice conjecture but Xsan is designed to work with the current Xserve RAID, and very likely the first generation Xserve RAID as well.

[Sharewaredemon]

blame motorola! :p

Hahaha, or IBM

[~Shard~]

hmm

Could you make it any more blatantly obvious that all you're trying to do is bump your post count and get the first post by posting a one-word, completely useless reply? What's the point of even having the first post?!? Thanks for coming out, now how about actually contributing to the discussion and offering up something intelligent to say... :rolleyes:

There are numerous reasons why the xSan might be delayed. It may simply be due to technical hardware issues, it could be due to software issues, or perhaps Apple decided it would be more advatageous to launch it at MWSF, possibly with the new xServes. I'm sure the xSan will be backwards compatible with the old ones regardless, so it's not like Apple needs to wait until the new ones are out to release the xSan, but that could be part of Apple's strategy - perhaps to better market it and get it the attention it deserves, Apple is choosing to wait. I could see this being a good move, actually, as I'm not sure if many companies out there will fully realize the true and impressive potential that this solution will provide. Perhaps showcasing it at an event such as MWSF will help make te public more aware of its capabilities.

I'm defintely excited to see it, so let's hope it is released soon! :cool:

[sjl]

As for as earlier comments about this delay being related to a hardware release, that's nice conjecture but Xsan is designed to work with the current Xserve RAID, and very likely the first generation Xserve RAID as well.

Xsan is designed to work with fibre channel systems. Period. End of story. Whether the fibre channel drives at the other end are Xserve RAID first gen, current Xserve RAID, next gen Xserve RAID, IBM's FAStT (or whatever they call it nowadays), HP, or some other company's hardware simply doesn't matter as far as Xsan is concerned. If it does matter, then that's an artificial restriction put in place by Apple.

Conjecture: I wonder if Apple will also allow Xsan to work with Firewire hard drives. Probably not, but it's technically perfectly possible...

[Rod Rod]

Xsan is designed to work with fibre channel systems. Period. End of story. Whether the fibre channel drives at the other end are Xserve RAID first gen, current Xserve RAID, next gen Xserve RAID, IBM's FAStT (or whatever they call it nowadays), HP, or some other company's hardware simply doesn't matter as far as Xsan is concerned. If it does matter, then that's an artificial restriction put in place by Apple.

Conjecture: I wonder if Apple will also allow Xsan to work with Firewire hard drives. Probably not, but it's technically perfectly possible...

Hopefully you're right and Xsan will work with non-Apple fibre channel drives, in the same way that Airport cards work with non-Apple wireless access points. However, everything on the Xsan site so far indicates it's Xserve RAID-only.
http://www.apple.com/xsan/specs.html

A non-Apple element is required of course, in the form of a qualified (http://www.apple.com/xsan/compatibility.html) fibre channel switch.

[AidenShaw]

Conjecture: I wonder if Apple will also allow Xsan to work with Firewire hard drives. Probably not, but it's technically perfectly possible...

It might not be possible, at least not without a lot of work. There are a number of arcane commands in the full SCSI spec that are needed to coordinate a cluster file system on Fibre Channel or a shared SCSI bus. ("SCSI reservation" for example)

While 1394 is superficially a shared bus like Fibre Channel, these oddball commands would have to be implemented correctly on the SBP-2 protocol running on 1394 protocol to IDE drives behind 1394 to IDE bridges. The probability that everything would work exactly according to spec is pretty low - especially since many of the arcane commands are only needed for clustering. (And anything that's not exactly to spec would result in the loss of data....)

Here's a brief page that describes some of the issues for the Linux implementation of SBP... http://www.linux1394.org/sbp2.php

[visor]

All I wonder about is. why announce things that early in the development circle?
I guess the only reason is to get seeding and Beta testers - very wealthy beta testers on the line to get feedback.

[kiwi-in-uk]

- snip - why announce things that early in the development circle?
I guess the only reason is to get seeding and Beta testers ...

Yes that ... and also to "hook" potential buyers, early in their decision processes.

IBM is (or was) an expert at this technique, especially in the 70s and 80s.

Microsoft uses the same technique - crudely disguised as "information for developers", but aimed at decision influencers (i.e. Microsoft qualified technicians) in large organisations. Very effective.

[ASP272]

I guarantee you they are waiting for good 'ol Stevie to announce it with the upgrades for the Xserve at his keynote. We'll just have to wait and see.

[Rod Rod]

I guarantee you they are waiting for good 'ol Stevie to announce it with the upgrades for the Xserve at his keynote. We'll just have to wait and see.

If what you guarantee doesn't happen, what are you willing to do? It has to be something good for the guarantee to be worth anything.

Anyhow I disagree with the idea (offered by ~Shard~) that this will be introduced at MWSF to give greater impact. It's a delay, an announced delay, and a delay that's made the tech press. Delays are bad.

Xsan isn't the sort of thing the public gets excited about. iPods, iMacs and 3GHz G5s are what the public notices.

Come January 11th, we're more likely to see an apology from Mr. Jobs, or short of that an acknowledgment (similar to the 3GHz explanation), than an Xsan "now shipping" announcement.

[AidenShaw]

Yes that ... and also to "hook" potential buyers, early in their decision processes.

IBM is (or was) an expert at this technique, especially in the 70s and 80s.

Microsoft uses the same technique - crudely disguised as "information for developers", but aimed at decision influencers (i.e. Microsoft qualified technicians) in large organisations. Very effective.

Then why is Apple starting lawsuits against the rumors that only help this process along?

Yes, you're right. "Asteroid" was supposed to be The Ego's "one more thing" - and that's been blown.

[Rod Rod]

Then why is Apple starting lawsuits against the rumors that only help this process along?

Yes, you're right. "Asteroid" was supposed to be The Ego's "one more thing" - and that's been blown.

kiwi-in-uk's comment was germane to the business market. "Asteroid" is for the consumer market. The approach is different.

Maybe the damage is done and M-Audio (the most likely candidate to do such a thing) is racing its own low-cost FW breakout box to the market.

[Jaz]

I'm annoyed by this. Nothing big but it is an option for consideration in my company for footage handling in FCP and the rest of our pipeline when we expand.

Apple sometimes resorts to the "when it's done" method of customer communication when they have nothing of substance to say. Consistency of message would be nice.

[cnoellert]

Hopefully you're right and Xsan will work with non-Apple fibre channel drives, in the same way that Airport cards work with non-Apple wireless access points. However, everything on the Xsan site so far indicates it's Xserve RAID-only.
http://www.apple.com/xsan/specs.html

A non-Apple element is required of course, in the form of a qualified (http://www.apple.com/xsan/compatibility.html) fibre channel switch.

It does and will work with 3rd party FC devices. What has been problematic is Apple's support for wwpn. In short order it basically means that it's one lun per controller in single controller raids.