Posts by M. B. 263 publicly visible posts • joined 26 Aug 2009
...they are able to buy the kit they think they need, it's not exactly newsworthy, is it? That kind of hardware would amount to a PoC lab or side project for large enterprises. We have a customer who bought 16 loaded UCS chassis and a couple shmill in switches, can I submit an article about them?
As a past customer, I've seen Dell take a big dump on their competitors pricing on many occasions and win business as a result. They won a lot of mine for offering the features I want at lower prices (with good local partners which helps a lot).
My only concern is that with reduced margins comes reduced R&D spending. Although as a private company, they can dictate their margins with no regard for Wall Street. We'll see how the strategy plays out over the course of the year.
Cisco is actually quite big in the Openstack field and uses Inktank on RHEL for storage as it is, so they don't necessarily need VMware at all. They are also ensuring the latest generation of products has full Hyper-V interop as well.
I am more expecting a Red Hat buy from Cisco though (now that Red Hat owns Inktank), that would have significant results on the quality of their Linux-based work.
The Compellent flash stuff works really well, having owned a hybrid systems with flash and disk. I found I had to let the data settle when migrating to the array so as to not fill the SLC tier too quickly. Once the colder data has de-staged to lower tiers you just let all the new writes hit SLC and let the array sort out the placement as it cools. Typical flash array performance, sub-millisecond latency and ridiculous IOPS, plus some added efficiency through thin writes and de-staging to more cost-effective media.
I'm actually not a customer anymore (I resell HP, EMC, and Nimble primarily and there's plenty of bug notices to go around), but I was a customer so felt the pain firsthand when a NetApp originally configured as a gateway decided to ignore all of it's natively connected shelves after an ONTAP upgrade. A DS4300 Turbo that would reboot randomly due to a bug in a watchdog timer. A remote InForm OS update which hung an F200 system, requiring a long drive and manual power down. An EqualLogic that wouldn't boot at all after a firmware upgrade, and then ran at half performance for two weeks before it got fixed. A nice, big V7000 system that could barely manage 20MB/s for iSCSI after an upgrade (I can assure you, that was plenty "disruptive" for my users).
Now when I look at the list above, that's a midrange NetApp, an Engenio-based IBM, a 3PAR, an EqualLogic, and a Storwize V7000. None of those are bad systems. I literally bet business on them being "not shit", but still there were bugs and odd behaviors. That is simply expected. It's why I always backed everything up to my Data Domain, then ran a clone job to send it all to my Spectra library before "non-disruptive" upgrades.
Now I DID miss the part about this being a disruptive upgrade (my bad), and that is definitely unfortunate. There's degrees of disruptiveness, from simply offlining both controllers at the same time to requiring complete evacuation. I would be pretty pissed about that myself especially if I had nowhere else to put my data and I absolutely had to, risking data integrity otherwise. I'd even be pissed if I just had to shut everything down to reboot a controller pair.
...someone blogged something about a rumored problem with an upgrade that's still in beta? I'd be a whole hell of a lot more concerned if it was production code, but according to EMC this is not the case. Sounds like a bunch of FUD to me (at least until XtremIO's start wrecking client data in the field post-upgrade).
For the record, I've seen just about everything shit the bed at least once during a supposed "non-disruptive" upgrade. Good backups are everything. Snapshots don't count.
I was concerned as well when I saw this. I had a couple "failed login attempts" from somewhere in northern Nevada a few months back and I used Keypass to generate a strong random password and enabled two-factor. Also did same for Windows Live and Apple ID, and proceeded to generate random Keypass passwords for any sites using my Gmail account as a login. Now I just want my financial institutions to all offer the same.
...since I have a personal HTC One and a work iPhone 5, and when I compare the two I think to myself I do prefer the iPhone 5 in daily use even with it's lower PPI screen, but I like the size of the HTC for reading docs a bit easier. It's basically exactly what I want, even though I'm not excited by it at all. Sigh.
...an agreement with Simplivity allows them the flexibility to lead with that option when it is the strongest option available without stepping on the toes of partners like the EMC federation, NetApp, and Nimble. Outright buying them only creates a quagmire of corporate bullshit between existing partners which I am sure they would rather like to avoid until they have a more accurate indication of how this agreement will pay off.
It's also worth noting that Cisco is an Inktank customer, running Ceph on UCS nodes for OpenStack storage. If they really wanted to make a move to shake things up and piss a lot of people off, they would buy Red Hat (who now owns Inktank), landing them an OS and hypervisor company with a scale-out disk storage offering to pair with their flash accelerators and a ton of open source know-how and resources to do pretty much whatever they want, extending the expertise into all of their existing products as well.
I ran into that too at a previous job, Oracle Linux on OracleVM results in cost savings that are hard to ignore. Couple that with Windows Datacenter licensing and ECI including System Center and the advancements in Hyper-V and I'm starting to see a number of companies start splitting their virtualization infrastructure into separate silos to save money.
IBMs old servers, EMCs old storage. Wonder when they'll start reselling Cisco 3560's and Brocade 200E's?
That said, the 5100 was pretty popular with telcos in distributed offices. I saw a bunch of them in that role and as far as I know, they are still there. Most people buying them for general usage were pushed towards the 5300 from what I've been told so they could have the option for more FAST cache and file services down the road.
...it depends on where you are located. As a customer of Dell, IBM, and HP, my preference was to deal with Dell ProSupport above the other two. I never got someone in a "faraway land" and issues were always resolved quickly and with less hassle. HP was easily second so long as you got a decent level one tech with a... less noticeable accent. IBM was always a gong show (how many times do I need to speak to fulfillment on a single issue?), no matter who you were speaking with.
As a partner who leads with HP and provides the break-fix in our geographic location, I can tell you our customers first level experience results in regular complaints to our account managers but second level and on-site guys are easily meeting the SLA requirements and doing a good job overall.
...at how well they are doing after taking a close-up under-the-covers look at their tech. Nimble came in and put one through the ringer for us and the experienced storage guys walked away quite impressed. That CS210 screams, even under the very unfavorable conditions we subjected it to. Not bad for entry-level. The economics hit a sweet spot as well, but they will run into a problem if they don't turn a profit eventually.
Still curious if/when someone going to make the first move towards buying one of these start-ups.
Nice. There are 6 of us going through NIOP training tomorrow and Nimble is selling exceedingly well for us - almost every storage discussion becomes a Nimble discussion at some point, if only for a few minutes. We have a number of customers who either want to stick with FC after committing to fabric upgrades and a few others who are growing in terms of cold data (IOPS don't vary much but the capacity requirements would put them at or over the limits of the current arrays).
...to see the inclusion of Fiber Channel HBAs. I've heard "it's coming soon", but that's about it. At least the midrange stuff now has additional front-end ports. I'd still like to see them scale up higher in terms of disk shelves, seems dumb to arbitrarily limit their capacity to 3 expansion enclosures when there's nothing really limiting them architecturally.
We have several customers taking a cloud-first approach to core infrastructure services, it depends on the level of confidence they have in their service providers (including us). Some of them are so cloud-focused they are even buying cloud-managed networking/wireless/security and getting someone else to manage that for them as well. We've got a lot of businesses on our hosted Lync service and have had great results and growth in that area. Putting VoIP in the cloud is a big commitment, but it seems to be working out well.
A lot of it boils down to cost. I can put 10 medium instances of Windows Server or RHEL in our cloud for 5 years for less than the cost of a single server with local disks (never mind licensing etc). Since most of our smaller customers are running less than 10 distinct workloads (and simply using hosted services for common workloads like email and the aforementioned collaboration), many of them don't even bother with a server at all.
...as well, since we're a premier partner. We seem to be losing a lot of disk-based business when we lead with HP. It's not that the deals aren't there, we just aren't winning them with HP. I'm not sure if it's a mindset thing or what. Might be a couple good quarters around the corner, but the past two have been pretty bad. Even traditional all-HP shops are taking the time to shop around and it hasn't been good for our high-end storage practice (and devastating on the low margin SME stuff).
There are at least a few VMware SE's mentioning that to customers scoping storage as well. Since we don't sell NetApp or HDS and EQL is our number 2 or 3 option for iSCSI, we're biting it occasionally when that question comes up. I swear there's another partner out there sabotaging us (good for them, I'd do it too). Other vendors promises are just that and most customers with any sense stopped believing those a long time ago (even though we lead with them usually).
...most partners aren't even allowed to sell NSX services or support yet. We're a large regional Premier partner and we're only just starting to get our partner briefings and there is some talk of training plans in the next 6 months. We're considering coupling NSX with Brocade VCS for a large data center build but yet we still have to wait our turn. Only select nationals with PSAs are actually permitted to sell the NSX product and services around it. So most people who say they've worked with it are full of shit, at least around these parts.
VMware is playing this really close to their chest right now, way too close to tell what impact this is going to have on the industry. Everything is pure speculation at this point. There is one large client that was looking for a significant amount of work to be done on their freshly implemented NSX environment around scripting and monitoring that is currently unfilled simply because no one knows the product yet.
@Tokoloshe: We're expecting that, based on what our PSE discussed. Some things are best left to ASICs, and when we pushed for more details around that statement and the impact of extensive ACLs and routing and load balancing configurations we didn't get very far.
I think there are some advantages to the per-core billing model.
I worked for a company that used the Oracle Database Appliance to drive a RAC cluster. It was pretty simple for me to take a look in my storage management tool and server monitoring tool to turn around and tell Oracle exactly what size and amount of disk I/O and CPU/RAM utilization we were driving.
Their proposal had us running Oracle VM on the ODA hosts and running the RAC nodes as VMs on the hosts (which is fully supported and gets around their restrictive virtualization licensing). It ended up being a significant savings for the company and upgrades were dead simple (just add resources to the VMs as needed). Since Oracle provides Oracle VM appliances for many of their applications, provisioning new applications was a snap.
YMMV of course. Worked well for that company though.
...there will be any encyclopedias or scientific periodicals in the huge collection (and if they will be searchable). I hate it when my son has a project and my wife's first instinct is to Google something and trust the contents in the first hit are factually accurate and not subject to bias. It would be nice to get him started down the path of proper research and use quality references that don't start with a "W".
Technical books, especially certification books, likely will not make the list since they are a good cash cow for their prospective companies but if they do make the list I will go straight to Amazon as soon as it's announced and buy a Kindle PaperWhite and a subscription. Fingers crossed on that.
"The Compellent array maxed out at around 6,000 IOPS but the Tegile Zebi hit 35,000, meaning more servers and users could be supported."
...sounds like someone either undersized or incorrectly configured the Compellent array, since a pair of SC8000 controllers with 64GB of RAM can do 6,000 8k (avg) random IOPS with 70%-ish reads on 24 drives - I know because I did it myself and hit 7422@10.25ms response time. A Compellent will not "max out" at 6,000 IOPS, not even close.
For reference, the 6 SLC + 6 MLC flash/hybrid shelf is designed to sustain 77,000 IOPS with sub-millisecond latency.
So either someone screwed the pooch designing or installing, or alternately, someone is lying.
We have a couple 3PAR customers running Tier 1 workloads on their mesh-active 10k systems and know many large local orgs using cluster-mode NetApp to run their Tier 1 workloads. Any by Tier 1, we're talking about utilities and hospitals and governments running hundreds to thousands of critical applications. Even with hardware failures or under extreme load they've all been fine (as long as they've been implemented properly, everything sucks with bad design).
I've been considering them both Tier 1 for a while now. Drives our EMC partner SE mad.
...much about this company except one partner who went with an ISE box over a P4000 solution we priced out. They went through 2 of them and hours of support calls trying to get the thing working properly. Still not sure if they got things sorted out or not. To be fair, that was a few years ago (3-ish).
...most of the deals I've been in leading with Cisco UCS have still gone to EMC or Nimble when it comes to storage, depending on customer requirements. Cisco account managers have only put Invicta in for special use cases.
They have been very aggressive with discounts for the VSAN nodes though and I too have heard about the Simplivity OEM talks/rumors. That combo plus ScaleIO through their BFFs at EMC would probably scratch any hyper-converged itch they have for the time being.
I still don't think they REALLY want to get too far down the path of being a distinct general purpose storage vendor. What I've seen locally is that a company buying Dell servers will buy Dell or EMC storage (or NetApp through someone else). HP buyers will buy HP or EMC storage (again, sometimes NetApp). Cisco UCS customers feel free to buy whatever they want, and we just try to encourage them to make it something we sell (all of the above, plus Nimble but not NetApp obviously). Cisco is happy because people jump into the Nexus product line as a result, they don't seem to care who wins the storage business.
I suppose we'll see. UCS is eating up the large enterprise/utility/government/cloud services market here right now but these places are usually the Cisco-or-nothing types anyways (at least on networking/wireless/collab), and even then I can't see then jumping storage platforms.
I would nickname it the c*ntblock though.
...is why I selected RHEL for a DMZ project involving a few dozen servers at my old employer. Sitting outside the managed environment meant sitting outside the visibility of most of our tools. No BS from Red Hat (here's how you license it, yes that's production support, yes that includes all this extra software, sure we'll give you a discount on JBoss), and the support was at the least slightly better than MS, and quite often much better.
The next year saw us replace aging Solaris boxes with RHEL on x86 for Oracle RAC, inexpensive, fully supported, and fast enough for our accounting folks to bring us cookies the first time they ran the monthly reporting.
My director was especially tickled by the amount of money NOT given to Oracle and Microsoft and the users were much better off for it. The yearly cost savings in licensing alone helped them justify two additional FTEs plus training for all my admins.
"Dell could have integrated its own server, storage and networking kit but has decided that it is better to use Nutanix as the software glue for its converged system offering."
Except Dell did integrate their own server/storage/networking kit with custom management software and they called it the Active System, orchestrated by the Active System Manager software. They still sell them, and they are owning the education market pretty hard in my area (as an HP and VCE/VPEX partner, this makes us sadface).
LOL
But also, I'm interested in this one myself. I'm torn between this and a Lenovo Yoga as an "alternate device". I like having the tablet functionality with touch for drawing network diagrams or reading manuals (or watching Netflix), and being able to convert to something with a keyboard when I need to create documentation or enter commands into a CLI.
As it is, I'm using an HP Elitebook and iPad Mini to get things done and it's a bit cumbersome with two devices, plus smartphone.
...but I don't think they really expect it to specifically protect them from EMC where EMC has provided evidence that data was removed from their internal systems by staffers who then left for Pure? That is not a matter of IP infringement, but rather a matter of IP theft (not yet proven in court AFAIK).
I don't think this article is telling the full story, as indicated by Neo Darwin above. IBM is saying something slightly different than Pure about the patents (licensed to Pure, rather than owned by Pure). Pure may have been forced to do so by IBM after violating a patent, but is now protected by IBM's patent muscle in case anyone decides to go after Pure.
There is a special layer of hell reserved for patent lawyers.
...or not having Fibre Channel is irrelevant.
It doesn't matter how you connect to the storage platform, so long as the "how" meets the latency and bandwidth requirements of the customer and is supported across the stack. FlexPods, for example, use NFS over 10GbE for storage connectivity. They are plenty fast. The baby vBlock systems use iSCSI for connectivity. They easily meet the needs for the workloads they are intended to support.
The Nutanix strength is also its weakness, it scales in fixed increments. If those increments are generally in line with the business needs, then it makes a lot of sense. If you are supporting multiple applications which scale in different ways, it does not. When you deploy a data warehousing application on your converged infrastructure and you quickly need to scale up storage capacity, it's easy to do on a traditional converged system. Not so easy on hyper-converged, especially if you do not require the additional cluster capacity from a CPU and RAM perspective.
I'm in love with the idea of Nutanix for other types of applications which scale more linearly like VDI. The easier something can be, the better. For more flexibility (supporting multiple widely varying applications), I would tend to prefer more traditional converged platforms which allow for more granular scalability.
...is the reason they are tearing it up around here. If they are selling at a loss, then they need to be very explicit with customer as to what their plans are for profit. Growth is good, but if you aren't profitable there will be no money to reinvest in R&D and the product will stagnate.
That said, we resell HP and EMC as well as Nimble, and Nimble is outselling them both, combined, as of late.
A Cisco buy would have been a huge boost to their business, but the growing relationship between Cisco and Red Hat (after their Inktank acquisition) means they could deploy scale-out Ceph nodes on their C240 servers connected to a Nexus fabric, and leverage Invicta for flash acceleration. I don't know if Cisco really needs them any more. I guess we'll see if they decide to buy either of them (or both, that would piss off a lot of people and be hilarious for the industry in general).
...this comes to fruition, even as a check-box item it can only help them. More so if they can leverage their ASIC technology to accelerate the dedupe process, if it's as transparent (and effective) as the thin suite it'll be solid.
I wonder how a 2-node 7450 would then compare to a Pure FlashArray, considering the robust options and maturity of the InServ software and the availability of the 960GB SSD from HP (compared with Pure's 512GB modules).
What I saw specifically was IBM shops looking for a solution that IBM couldn't provide themselves - the N-series stuff is way ahead of the DS-series stuff on features and functionality (and was often put in front of existing DS storage to extend ONTAP features to the legacy arrays). IBM would pitch N-series to those customers to keep them in the IBM camp.
The downside for IBM is that IBM would often bring the first FAS into an environment, but once the customer saw the number of things they were missing as part of being an IBM customer rather than a NetApp customer, NetApp would sell the subsequent arrays (IBM didn't certify many versions of ONTAP and was often 6 months behind on software updates, as well products which became free for NetApp users were a pain to get a hold of for IBM customers, plus the My AutoSupport site was not available to IBM customers - big let-down, and further, the support was terrible off-hours).
I could see the deal being lucrative for IBM in the beginning, being able to offer a scalable, high-performance multi-protocol storage array without losing customers to competitors, but once those customers started turning to NetApp for future purchases the writing was on the wall for the OEM deal (NetApp specifically supported SnapMirror from N-series to FAS for data migration and provided the licensing free of charge for that purpose).
I don't know how the Unified offering works out for Storwize customers but I've heard mostly good things on the block side (especially for FCP, with a few odd comments about poor iSCSI performance under VMware but that was from a P4000 pusher).
...didn't do a very good job with the N-series. I had one and the support was miserable. 140TB in Production with top support and they wouldn't help with a failed ONTAP update after hours, hardware support only. They could never get the V-series license to work with their own DS4300 arrays (we had two of those as well). After a trio of massive failures (complete outages requiring the reboot of the entire storage stack) we were stuck with a bill for native disk shelves to replace the capacity of the DS4300's we couldn't use as advertised. Our account manager was kind enough to offer us a business card with "Sorry!" written on it.
The moment things started going to sh!t, IBM was in pushing the V7000, with no interest in trying to remediate the issues with the N-series. It was one of those things that you would have to remind salespeople of "You know you sell N-series, right?". The DS stuff would only get pushed to the smallest businesses, now with the V3700 pulling in the bottom end there is little point to continuing the relationship I suspect.
The brand recognition may have been nice (I was told IBM will sell the first one, then NetApp will sell all the rest) but I don't think they need it any more, to be honest. Not that it will help them, but I don't think they should be expecting much from IBM any more at this point.
...silly not to include the 3PAR 7450 and Compellent SLC/MLC AFA options simply because they are also sold with spinning disks in other guises. HP and Dell have dumped cash into optimizing those platforms for all-flash and they are marketed as being all flash options. So is the VNX-F 7600, come to think of it. They also offer a bit of a value proposition for HP/Dell/EMC existing customers around training, and many of the advanced array features some of the flash vendors are missing are already available to these models which are built on a known base product. I've used them all in their spinning disk forms and they're not lacking for features or performance capabilities.
I think they offer a ton of value in the AFA market, even if they are just flash-optimized versions of rust-spinners.
Out of curiosity I asked my contact who went down the Pure route how the whole thing turned out since I heard the "cheap disks" thing second hand.
Turns out Pure threw an agreement together whereby they would renew their support early and co-term the support for an additional shelf, and they'd get the shelf free because Pure screwed up. So they didn't replace anything but they did get a big spiff out of the deal.
Might have had something to do with their "fill in" array being a Nimble though (they are also very happy with it, lucky bastards).
...a situation where the system didn't nearly meet the marketing claims as far as the storage efficiency numbers were concerned. They were told 30x which got a good laugh out of everyone at first. After "data modelling" Pure said no less than 5.5x (not that they put it in writing though).
Actual result was 2.3x and not enough capacity in the array as it was originally sized, so the customer had to buy something more cost-effective to round out the capacity deficit (they needed about double the storage efficiency number to ingest everything with a bit of breathing room).
Unfortunate and very disappointing. My God is it fast, though! The CIO was pissed off and seeing red all they way up until all the comments on how amazing the ERP and CRM systems were running started to flow in.
I do believe that Pure ended up upgrading them to newer shelves at some point for dirt cheap dollars though. Never heard how that turned out.
Wow, imagine that!
Seriously, that's been a missing part of the formula for a while. I believe the technology existed in the DS8k series for a while but was sorely needed in the midrange. As soon as EMC, HP, or Dell heard storage tiering was a consideration in an RFP they would shred the Storwize on the lack of 3 tiers.
The thing of it was all 3 were doing it and claiming success, meanwhile only IBM was doing 2 tiers and claiming similar capacity efficiency through real time compression on fast disks. That made IBM the odd-man-out in competitive bids where tiering was important and I know it cost them business locally.
...34% of the array total cost from one vendor is in just software licenses/features (that are advertised in a manner which would indicate it is included with the base), that's the average of three customer quotes for the same model of device.
Vendor 2 works out to roughly 32% of the total array cost in just software bundles.
That's terrible, they should feel bad.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
