Thread regarding Oracle Corp. layoffs

Who is "we"? Why is Marketing trying to convince a board for people laid off/going to be laid off from Oracle how great SPARC or Solaris are?

No one here is going to buy Oracle–anything, let alone these relics from the past.

Marketing really needs to rethink their strategy. This post actually reinforces the fact the SPARC/Solaris is slowly sinking.

I run Oracle Solaris 11.4 at home as a full e–mail server. Never wanted to touch the idea of moving to linux as zfs, smf, dtrace, altboot features cant be beat. However, in the enterprise – that's a different story:

1. What doesn't Oracle Cloud even offer Solaris x86 as a VM?? Common!
2. OEL versus Solaris is a tricky one with Ksplice. Still, rather have my heart monitor hooked up to a solaris machine than linux ;)

Going away now

can't believe people are still talking about sparc or solaris.... Oracle canned the development years ago, now also Fujitsu, the last one still doing something on sparc, has terminated all development, see the latest public roadmap where all the new sparc processors have been cancelled.
sparc is dead, solaris is dead. get over it.

@4tkuf+YVfhyrC:

Are you high?

My bet is that you're one of those SPARC old-timers who got canned in 2017 and still hasn't gotten over it.

It's always a good idea to check in with reality once in a while.

@4tkuf+YVfhyrC

Stop deluding yourself. If the combo of sparc and solaris are so wonderful, why are they both dead? No roadmaps, no new releases, etc... About the only customers left are ones running legacy stuff and biding their time until O k–ls the support. No one in their right mind is going to deploy solaris and sparc on anything else.

I see all these id–tic comments from Linux people who think Solaris is an old or aging OS..

YAWN... WHATZUP BOOMER?

I see all these id–tic comments from Linux people who think Solaris is an old or aging OS..
I have stop and wonder what these people actually do with their Linux OSs?
Let's start with features.
Virtualization ? Solaris comes complete with virtualization built into the OS with either sparse zones, full zones or Kernel zones. You also get the additional layer of LDOMs, which can stack methods of virtualization. Linux offers – {crickets}
OS Licensing? Solaris maintenance is free with the purchase of hardware. Hardware, I might add, that is not much more than competitors.
Filesystem layouts? Solaris comes with ZFS - the fastest and most secure filesystem structure around. It also offers the ability to mirror, clone, snapshot and much more. All this for free. Linux includes.. ... lvm? HAHA.. speaking of the 80s.. they want their volume manager back..
Let me include some extras that I know Linux doesn't offer-
dtrace and the ability to traverse what's running in the kernel.
etherstubs
Alternate boot environments that can be mounted, manipulated and patched against .
fast reboot
More secure
The packaging system that prevents ever needing a major OS upgrade (you know, like rebuilding your host to go from 6.3 to 7.1)..
SMF - Service management framework (you know, that think Linux is trying to do now)
Remote host administration
Free virtualization management - opsCenter- which allows for live migrations of solaris hosts and one button press OS and/or firmware updates on as many hosts as you'd like.
automatic/reactive networking -
More usable IP stack using dladm and ipadm command structure
Bridging - supports both STP and RSTP -
Integrated Load Balancer provides L3 and L5 load balancing capabilities
and that's just sc-aping the surface. Solaris is better than Linux in every way.. So much so that they slowed development to save money.

Now, as far as Sparc's superiority over all other hardware platforms.. That's even more fun -
Threads per socket - 256 - x86 (not fair to even compare, since hyperthreads aren't real threads)
Built in database accelerators.
Built in Compression accelerators
Built in encryption accelerators
Reduced Oracle database license costs - .5 core multiplier coupled with the true removal of threads - So a 256 thread box can have a single zone or lpar using 16 threads total and that will be the equivalent of 1 CPU core for database licensing.
Their M series CPUs are so far ahead of the pack that they stopped development on them 2 years ago and the rest are still working to catch up ..
1 socket gets you 32 cores with 8 threads per core clocked at 5Ghz.. Meanwhile, Intel is squeezing blood out of their turnip dies and getting up to a whole 26 cores with c-ap hyperthreads on that. If I have to explain the difference between multithreading and hyperthreading you don't deserve to be a part of this conversation.
Then there is reliability. .. Unmatched by any of the x86 vendors in terms of total system average uptime - You can't compete with the hardware that Oracle produces.

sure, mainframes are irrelevant to IBM, as much as 10% (+ another 10% for services) of total revenues are irrelevant to Oracle.

Yeah and people still spend money on mainframes and COBOL too, and coal power plants still exist. That doesn't equal relevance.

"hmmm- if you're referring to the onprem support revenue,"

hmmm.... what's unclear in the phrase "the same people spending 2B$ on new SPARC systems every year"?

It seems quite clear to me, but I'll try to rephrase: "the same customers that still buy new SPARC hardware (T8 and M8 servers + Fujitsu's M12 outside of Japan) and still spend more than 2B$ per year (as of FY18 financials, FY19 numbers are little bit less but still have to be consolidated with Q4)"

Financials are public, you can check by yourself. "Hardware" revenues are around 4B/Y. Exadata and other x86 systems count for more or less 40% of the total. ZFS, tape and other hw are almost irrelevant. The remaining is SPARC. By the way, this does not include Fujitsu SPARC revenues in Japan, which are quite large.

The above does not include services, which counts for another 4B/Y (as of FY18) but I don't have the ratio SPARC/otherHW so I cannot comment on this.

Stop chasing this dumb topic please. Some old folks are just waiting for their retirement. If you are irrelevant, go away.

"Who uses Solaris?"

"the same people spending 2B$ on new SPARC system every year"

hmmm- if you're referring to the onprem support revenue, it's for renewals for the support contracts, not for new sales of hardware. And the support is as bad as ever but they keep you on renewals pretending that bug fixes will be available - NOT!!! Hardly any engineering staff remain that have knowledge of the HW. It's sustaining support only and very little of it.

"Who uses Solaris?"

the same people spending 2B$ on new SPARC system every year

"...but you're irrelevant."

I would really really really love to be 2B$ so irrelevant, unfortunately my salary is much less.

If you think Solaris is relevant I hate to tell you this, but you're irrelevant.

Who uses Solaris? Nobody cares it's 11.5 or 12. Solaris management and dumb architects have been making wrong decisions for years and k--led the product. Time to move on.

Layoff off the magic mushrooms, sparc is not coming back.

"Remember when Sun's marketing campaign was "We're the dot in . com"? "

that was a marketing way to advertise the real Sun founding slogan "The network is the computer", precursor to the cloud. And in fact Sun was the first company to ever start a public cloud with Sun Cloud

... and then LE said what he said about cloud.

I like Linux red hat

@1wnm+YVfhyrC
Remember when Sun's marketing campaign was "We're the dot in . com"?

• and then the dot com bubble burst.

" but Oracle is no more interested because margin is too low even if much better than any other x86 vendor, and better also than IBM. "

All the more reason to unload it, it would seem. If it doesn't fit into your plans, and you can get a few billion for it now, as opposed to chicken feed later, why not let someone take it off your hands? Of course the only possible candidate is Fujitsu, and maybe they don't want to pay a lot for it. On the other hand the dismantling happening in Solaris does leave Fujitsu in a precarious state.

"I predict in the next few months Oracle will announce the release of Solaris 12."

not only we will not present Solaris 12, but also the announced Solaris 11.5 due out in july/august this year per the roadmap presented many times by BN org will not see the light. for sure not this year, most probably never. we actually don't have enough resources to present any major revision, only patches and updated FOSS.

Oracle SPARC is still doing around 2B/Y in HW sales, with an op exp somewhere between 0.7B/Y and 0.75B/Y (only SPARC HW, excluding cloud systems and support, including sales, marketing and R&D. almost all sales op exp, marketing and R&D are now almost zero), so yes it's still a profitable business, but Oracle is no more interested because margin is too low even if much better than any other x86 vendor, and better also than IBM.

I don't have support numbers, but I assume something better than the above.

So even if I truly believe Fujitsu is interested and committed in SPARC, mainly given the japanese multi-M$ market, unfortunately this collides with Oracle decision to terminate both SPARC and Solaris.

I'm not aware of any IP that Oracle has sold to a 3PP in the past, I would love if Fujitsu buy Solaris but I highly doubt it will ever happen.

I luv Linux, Solaris bites da big one

They will sell when Fujitsu agrees to pay them what they want for it, but right now they are still making money from it so they're in no hurry to give it away.

Yeah, as long as it's a cash cow, LE will hang onto the old Sun org. But as others have said, he might sell it, for the right price. The money would come in handy for stock buy backs.

The 80s called, and they want their computer back

"LE and MH walked away from negotiations with FJ about Solaris ownership a few years ago. So it was considered at one time."

They know that Fujitsu NEEDS Solaris if they want to keep their Sparc business alive. They will sell when Fujitsu agrees to pay them what they want for it, but right now they are still making money from it so they're in no hurry to give it away.

LE and MH walked away from negotiations with FJ about Solaris ownership a few years ago. So it was considered at one time.

"If there's any future for SPARC, it's with Fujitsu, not Oracle."

Totally agree but SPARC cannot have a future without Solaris.<<

Makes one wonder if LE will sell what's left of Sun to Fujitsu and use the money for more stock buybacks.

"If there's any future for SPARC, it's with Fujitsu, not Oracle."

Totally agree but SPARC cannot have a future without Solaris.

@sfn+YVfhyrC
I still remember the day they pulled OpenSolaris. That was the day Solaris starting dying its slow lingering death.

If there's any future for SPARC, it's with Fujitsu, not Oracle.

As for Solaris, it will cling to life support as Oracle won't just give it away (like OpenOffice, NetBeans, etc.) any time soon. Maybe when all the current contracts and commitments finally expire, but not now.

Solaris 12 was cancelled in late 2016. Parts were sc-apped, like Docker support, the rest turned into 11.4. When Oracle extended Solaris 11 support to 2034 it was to cover the decision to never ship a Solaris 12 release - if there was going to be a Solaris 12, there’s no way you get 15 more years for Solaris 11.

I hope what you say is true.

I predict in the next few months Oracle will announce the release of Solaris 12.