32 posts • joined 1 Feb 2011
RDMs and Snapshots
OK I'm going to ask the question here.
A stack of experienced VMware guys (above) don't see any problem with the 2TB VMDK limits. How do you do host-based backups and VM snapshots? Don't you need to have synchronicity between all of the vdisks? How do you ensure the SAN admin doesn't nuke the RDM snap before the VM snap (if that's even possible)?
And you'd lose Storage vMotion as well, right? Isn't that supposed to be a killer feature? It seems to lock VMs pretty thoroughly to the specific array.
Some of us still have physical servers you know
And yet, no sign of a replacement for P2V, removed from VMM 2012 R2. Which was removed, as I recall, so that it could be updated and released on an accelerated schedule.
Apparently the intended acceleration vector points backwards.
So I'll buy one for my datacenter and copy files to that filesystem with Explorer, right? What do you mean it's only in the Cloud? I want it local! I don't want to wait hours for data copies over this pathetic long distance WAN link! What about my backups (do NOT try to tell me that any form of resiliency = backup).
And what the dickens do you mean you can't just save to it with a shared drive - a REST API!? Oh, wait, no you want to sell me a NAS gateway too - for goodness sake, man, if it's a filesystem let it store files my way!
Stupid Cloud. It is NOT the be-all and end-all of IT (or if it is, I'm taking up flower arranging).
Re: Obviously not
Evan ... I think your calculator is broken.
32 x 96 = 3072 (if you prefer, 2^5 x (3 x 2^5) = 3 x 2^10 = 3 x 1024 = 3072
That's 3TiB in my book.
Well despite any disadvantages, there is one significant advantage to the MS-contributed designs - they fit in a standard 19" rack. So those companies with HP, Dell, IBM servers and blade chassis racked today in HP, Dell, APC and IBM racks can switch out those old servers for a "MS Blade Chassis" without also ripping out the racks themselves.
Yes, obviously there are potentially downsides (2 sleds per 45mm 1U, rather than [IIRC] 3 per 48mm OpenRU in the existing OCP designs) but at least it can be up to the customers which size to choose.
Of course, we won't know if the designs use the same or similar power interconnects, nor how the network and storage interconnects are handled, until the designs are available. But it certainly seems to me (and obviously Microsoft) that there could be value in sticking to existing standards for the rack rather than going the whole hog for the new OpenRack design.
Re: A NAS by any other name...
No it's not even a NAS - not really. After all a NAS generally exposes a filesystem onto which you place named files with SMB/NFS/etc. A NAS is ... useful! These don't even have that - unless you have no directories on your filesystem and you name all your files like this:
In which case yes, it's close to being a single-drive NAS.
It's more like a single-drive SAN, incompatible with the rest of the world. With at most 1GBps (that's 10Gbps!!) per device, the network ports will cost more than the drive itself. It's about 50-50 with a 1Gbps NIC on the device but then a drive will deliver 100MBps max (less than SATA). And you better hope the switch never fails, you'll temporarily lose access to data but not know what until you try to access it.
I'm sure this is a solution. But it seems someone forgot to investigate whether there was a problem in the first place.
Re: I couldnt agree more
OK I hear what you're saying, and yes:
Q. what is the first thing you do after you carve up a LUN ? ...
A. Put a filesystem on it
True enough, but I have a followup question. Which one?
Because the one thing I don't see addressed here is heterogenous environments. Which filesystem can my Win2012 box running Exchange share with a Win2012 box running SQL and a RHEL box running Oracle?
Or is everyone assuming the storage should be presented to a set of hosts running an intermediary layer and carved up there?
Funnily enough, though, most of my arguments with storage engineers have revolved around them wanting to give me little 3 and 5 disk RAID 5 sets for each Exchange database, or insisting that RAID 5 is better in every way than RAID 10 (spindle utilisation yes, performance .... not always despite what three-letter-vendor claimed). Not sure the LUN argument has even reached those cretins yet.
I just want to point out that a DNS blocklist as described (and though I do work for Telstra, I have no direct visibility of what's happening here) won't block sites that share an IP address with a C&C site.
As described the "filter" looks for the DNS query to badguy.domain.com and either blocks or ignores those queries. So when you look up "goodguy.mysite.com" it won't match the bad site DNS name, and your query (and connection attempt) proceeds.
I'm not a fan of filtering/blocking etc; be it whitelisting, blacklisting, or using a black box list of "stuff someone claimed was bad". But let's argue about the right stuff :)
LUN sizes - 2TB?
So for me the thing that stands out with all of the VNXe specs is the continued specification of 2TB maximum LUN size. In all of my (increasingly) meagre experience, the LUN is the unit of storage presented to the server by the array.
So what I obviously don't understand, in this day and age, why is 2TB still the maximum? MBR vs GPT for some reason? Or am I completely clueless (it wouldn't be the first time)?
I have SMB customers with larger shared datastores than 2TB - let alone larger enterprises (in which a 10TB fileserver is the small one in the branch office). Is there some unit of storage that aggregates 2TB LUNs together at the array level, or is one expected to present 5x 2TB LUNs to the server and aggregate them there (with mdadm / Storage Spaces / etc)?
Voluntarily submits to the courts!?
I liked this piece the best:
"Through discussions with the ACCC with a view to resolving the legal proceedings brought against HP, HP has voluntarily consented to Federal Court orders."
Voluntarily consented? Ah yes, the big international company (which IIRC has 10K staff in Aus) doesn't HAVE to comply with the Federal Court, it's just the easy way out this time.
IMO, the whole "apology" (like most, it's fair to say) is just another generic, self-serving, half-hearted, condescending, steaming pile of ... er ... codswallop, from a company who most certainly doesn't give two bits about anyone, or anything, other than screwing anyone and everyone it thinks it can while still getting away with it.
Erm, SATA is fine ... but not dual-ported
Actually SATA is supported with Storage Spaces, but it cannot be shared / multiple-connected like SAS can. Basically the drives need to be "dual-ported" (connected to both controllers). It's a small thing, but we like accuracy, don't we?
Storage Spaces then manages having the disks online on one node or t'other.
Which model is it?
The text says it's the DGS-3420-28TC but the picture seems to be DGS-3620-28TC - and both model numbers appear to be valid and quite similar?
Re: Free Is Good @Denarius
I didn't intend to say macro rewrites - if that's what you read then I apologise. The reality is though that templates written in Office 95 and Office 97 continue to work for the most part in Office 2010 and Office 2013. So that's a pretty good compat record. (And yes, there ARE things that break and those things are documented). Plus you get a minimum of 5 years support for Office (Office 2007 only just left support). How long is that version of LO/OO/SO (collectively $NEWOFFICE) going to be supported until you're told "Oh that's in $NEWVERSION, you'll have to upgrade". And how much is that pro support going to cost compared to the free MS cases that Enterprises get?
The problem is when you move to $NEWOFFICE it's not a tweak, it's a complete rewrite - assuming that the new product even has the right hooks and triggers for macros.
As for throwing away ... er ... replacing the DMS - let's look at that option.
1. Find a new multi-million $ DMS that works with $NEWOFFICE.
2. Write or port all functionality from $OLDDMS to $NEWDMS
3. Migrate all content, version info etc from $OLDDMS to $NEWDMS
4. Run both side by side for 10 years because no-one has the cojones to decomm the old one.
6. Rewrite macros and integration for each new version of $NEWOFFICE.
It's really not that much better is it - even assuming you can find this near-mythical $NEWDMS?
Re: Free Is Good
Because OpenOffice is a complete and perfect replacement for Microsoft Office, right? It never gets formatting wrong, understands all the existing 15 year old templates (across Word, Excel, PPT), includes a mail program people actually like (Outlook replacement) which talks Exchange, does offline caching and integrates with the email archive? And Excel formulas are all present with the same names, parameters and results?
And there's no chance LibreOffice or OpenOffice would be unable to create and edit documents using Rights Management Services? And it also works with the appropriate DMS (which is usually a fragile set of poorly written and undocumented macros which often needs updates even for Office service packs, let alone completely different products)?
Sorry, but while OO and LO are great for some people, they're not there yet for everybody.
Even Paris knows the above...
Try leaving them on for more than a day
I'm serious - I have a customer here in Sydney with a pair of earlier Dell switches. After about 10 weeks of uptime, something goes wonky and they stop forwarding packets at or near line rate (think 20kB/s for a GigE port). So now we have a monthly switch reboot for this customer.
Re: I'm wondering ...
I believe a magnet program is a program for more advanced learners - the idea being that you attract the smart and/or enthusiastic kids and the rest will follow. So it's also somewhat of a "comply or go back to the less advanced classes and be bored/lose opportunities".
Re: Vodafail anyone?
Vodafone has double the spectrum of either Optus or Telstra, as I recall. ISTR that the 700MHz spectrum was allocated 25% to each of Telstra, Optus, Voda and 3; then Voda bought 3 (and all the assets, including spectrum). Given the lacklustre performance and the resulting mass exit of subscribers, I don't see them being short on spectrum right now.
Re: 10GbE is still a bit TOO quick...
While that's absolutely true, there are a couple of factors that many (including I) thought would force adoption faster.
First and foremost, virtualisation. When you have 30 VMs on a single host (not hard to do whether you're ESX, Xen, KVM or Hyper-V) even a 4Gbps channel averages out at under 135Mbps per VM (and you're hoping the peaks on one VM cancel the troughs on another for higher throughput).
Secondly, iSCSI storage, where 4-8Gbps total from the array might be OK for smaller environments but isn't enough for larger ones, especially during backup;
Thirdly, server-backup. Aggregation only helps so much, as many (most?) aggregations split data across the connections based on source and/or destination IPs. So a single stream is limited to 1Gbps.
All those scenarios would fare better with 10Gbps, especially if all the vendors start doing the funky network/bandwidth splitting like HP - where each 10GbE can be logically separated into 4 different virtualised adapters for the OS.
In those cases, 2 x 10Gbps connections provides similar connectivity and throughput to 12 x 1Gbps connections. If 10Gb is only 5x the price, it's CHEAPER than doing 1Gbps.
Forgot about power
The other key advantage the older CPUs have over current brethren is the power required to operate them. For example the 386 required 400ma at 5V - just 2W - and the radiation-hardened version 460ma at 5V (2.5W at 100% usage). See this doc: http://datasheets.chipdb.org/SEI/space-elec-80386.pdf. And that's the 1995 version so it's probably less now.
Even the lowest-power Intel x64 chips require more than that (17W).
And apart from generating the power, in space we also have only radiation to disperse heat. No convection, evaporation, sublimation, etc; forced cooling with air or water just moves that heat somewhere else to be radiated.
Re: Bullsh1t not baffling MY brains.
Slightly OT I know, but try installing drivers for the laptops from other download sites (e.g. Aus or UK), or set the driver properties to a foreign country (again, Aus is a reasonable choice). Or tell the AP it's in Aus.
You should get the full range of 2.4GHz channels then.
Yeah - "normal" thing to do
They all do it. They claim they've selected certain drives and made them differently, in an "enterprise" ready state. Frankly I figure the extra bucks are twofold - one, to get enough dosh to cover the extra warranty claims from running the drives 24x7 for 5 years instead of 8x5 for 3 years (it's 44000 hours instead of 6280), and two, some extra margin for the resellers to make some money off the enterprise drives.
I really do doubt the quality is that different (of course having said that, I'll get home and I'll have lost 4 of the 20 disks in my array of consumer drives).
Sherlock would know whether or not the drives are different ...
Re: Who exactly does this kill?
I'm not sure what you're trying to say with "adding switch capacity with every enclosure" - since even with UCS, you need to add fabric modules to each new chassis. And you need to plan for resiliency there, too just as much as you would with a bog standard switch.
And looking at the Cisco modules, depending on your point of view, they're either switches themselves (just switching something other than plain Ethernet) or they're dumb devices for aggregation of bandwidth. One view has them adding switch capacity with every chassis, the other says they're a reasonably dumb and inefficient piece of circuitry for which Cisco charges as much as a small car.
How is this not extortion?
Regardless of whether the person has committed a crime (or breached the law in another way) or not, nothing has yet been proven.
Near as I can see from the article, and thinking of each individual case, there is a threat against the person, accompanied by a demand for money to make the threat go away.
Since the law firm is quite obviously a private concern, how is this not extortion (demanding money with menaces if you prefer)?
Re: had this one yesterday
The magic is in learning how to answer, not what the answer is. Users do not listen to explanations, because the computer is smart and should just know that they mean "go to www.google.com" when they randomly mash the keyboard with their fists.
In your case, the appropriate answer is that the application is already installed on all computers in the company, and the shortcut is under Favourites. If you wanted to get even more fancy about it, get something like WiX and create a little MSI that creates shortcuts to the web sites under Programs, and users won't even know the difference.
But the answer to "Please install web application X" for your users should always be "It's already installed", which solves the problem and makes you look like a genius for anticipating their problem.
Re: Experience first, then we'll talk
I'll have to disagree with you - that's kind of the point of Datacenter Edition. Purchasing the licenses for the host entitles you to run any number of VMs on that host without paying for other Windows OS licenses. So if you license the cluster with DC edition, you have as many VMs as you want and you can move your VMs around any time you please, because the license isn't on the VM it's on the host.
Now applications might be different, but you generally license the VM for the application not the hosts; in which case you may or may not be fine. But Windows is a no-brainer.
Experiences vary widely
This would be in direct contrast to my own Win8 personal deployment. It's a member of a domain, which doesn't have a large amount of group policy. SCCM client is fine. Office works fine. Metro apps launch just fine (your experience with Metro apps not launching has been noted by others on the MS forums). So it'll be a group policy breaking Metro apps, somehow.
I'm using it as the day-to-day host for a stack of VMs (including my corporate SOE VM) and running apps, connecting to other networks etc. I've not yet had any crashes or BSoDs, though VPN setup was annoying, requiring a logoff and logon to connect.
The inbuilt mail/calendar clients are ActiveSync clients, so the PC is treated the same as a mobile device (enforce PIN/password, screen lock etc). If you have working ActiveSync and AutoDiscover then the setup should find things; any problems will likely be that the internal and external addresses for servers will differ (so they may work only in the environment in which they were set up originally, be that internal to the network or outside the network).
Biggest annoyance is Metro apps not working behind an authenticating proxy. This is probably the killer bug for any corporate deployments - they HAVE to work, and work well. Right now they don't work at all.
If you've used the Win7 Start > search functionality, or you use a scroll mouse, the Start panel should "just work" - you hit the Windows key, start typing what you want, and press enter (maybe an arrow or two). Or you hit the Win key and use the mouse to scroll (though I still wish the Start panel would treat mouse-down the same as touch, and mouse-up as release on the background so you could drag it).
Overall mark: 8 out of 10.
Biggest takeaway: Stop moaning about the start panel because it isn't going to change. It's going to be there on Windows Server too. Try USING it for a full week - don't stop after an hour because you don't like it. And if you don't believe me, cast your mind back to 1995, and I pretty much quote the newsgroups: "OMG the Start button is horrible! I want Program Manager back. Why should I put my mouse all the way down to the corner to start a program? What's wrong with the way it was?"
@ShelLuser: He IS the network/server admin. He is the one who SHOULD be trying it out, though perhaps not the daily driver, even if that's also my preference.
Re: Uber fail of epic proportion
You said: "I couldn't stand using it for more than about half an hour. Just the fact that I couldn't pop up the Start menu and bash a few keystrokes to find EXACTLY what I wanted in seconds really annoyed me."
And with this you're just trolling. If you hit the start button, or the windows key, then type, do you know what happens? You get search, just like Win7. And it searches all the programs, settings and documents, just like Win7 - except they're now separated into Apps, Settings and Files groups.
Oh ... wait a minute. That's just like Win7.
Oh, I forgot something ...
How much of the cost of a copper phone line is the maintenance of the existing copper network? Given Telstra wanted (I think) $20 per month per line including margin ... that suggests $14 per month is a reasonable average cost for maintenance of the copper. Compare that to $6 per month for original amortisation, or ~$14 per month if you include interest, to install the fiber.
You can't be serious. Fiber in conduit, in the ground has a lifetime significantly exceeding 3 years - indeed, barring rats chewing on it, excess input power burning out the remote end (which is easily serviceable), thermal shock and the like, it's like that broken beer bottle on your lawn - it'll last almost unchanged forever, until you step on it.
Try 50 years as a useful life (lots of the copper is older than that) - remembering that trans-oceanic fibers are 20+ years old and not stuffed by 20 years of salt water immersion!
~3666 per household, amortised over 50 years is $6.11 per month (ignoring inflation/cost of money). Given that's what Telstra charge to display a damned phone number when you ring (caller ID is $6 a month, IIRC, for having a software option switched on instead of off - what a RORT), and we're building a national fiber network - I think that's a damned good price. If you prefer - make your sandwich for lunch instead of buying it twice a month and you're in front.
But wait ... this money is coming from taxes. Investment in infrastructure, by the government, during a recession (ooh that'll bring out the haters). I'm generally anti-Labor, but it's the right thing for the country. Oh and the children (that'll win the argument). Something about thinking of the children.
Your long term solution is ... not to put to fine a point on it ... unworkable.
Not only would all the credit card providers need to agree to collect tax on behalf of the Australian government, if it succeeded, every government would end up wanting their cut.
Furthermore, what tax rate should be paid and to whom, if the goods are bought from a UK store by an Australian CC in Australia and sent to a UK recipient? Does your answer change if the Aus CC is used in the UK? What about a UK CC being used in Australia on holidays to purchase and send the same goods?
What tax rates does Visa collect from an Australian shopping in the UK using a US Internet proxy for delivery to Germany, for a person on holiday from Brazil? As you can see the possibilities, and associated complexity, are endless.
Finally ... while I have the utmost respect for people who work retail (having done so myself years ago), people have no right to a specific job. And ignoring the direct financial aspects of it for a moment and dealing only with the underlying morals, why does a person in your own country deserve a high paying job more than a person in China deserves a lower paying job? Citizens of AU (and I suspect most 1st world countries) can get support when out of work - perhaps not so for others.
I think it's you doing the joking.
Yes, Internode is a premium provider. Still not as bad as Telstra (was?) though.
Where can you get a phone line + 25M/5M service (guaranteed sync rates) with 30GB of data for $70? Or 25/5M with 200GB for $90? I'm paying $90 now on a Telstra port for that much data and lower sync rates (and yes, I'm with Internode) - and that's not including the sodding copper (which we pay for and then Telstra attempt to ignore service and support). For me it's $25 cheaper with those published prices, and I suspect I'm not alone.
Heck - show me anyone who will guarantee 12M/1M sync with a phone line today at those prices (don't forget to drop the $25 off the price for the phone service)? Also note the word guarantee there.
Dodo (widely regarded as a "cheap" provider) lists their current 10GB DSL service as a total of $50 (or $35 as a Naked service) - 100GB (if you can get it, that is) is $10 more. So cheap provider of DSL of whatever speed the line is capable of, with dial tone for $20/m less than Internode. Good luck with the contention ratio though, there are no guarantees of sync rates let alone performance.
I'd say that the plans are actually pretty reasonable - especially for those of us who remember 500MB with Telstra Cable for $100+, and 19c/MB.
FTA: This Cougar Point SATA I/O flaw does not seem to be as serious as Intel's infamous June 1994 floating-point math flaw, discovered in its Pentium 5 and not confirmed to customers until five months later after news of the bug broke in the press. That bug caused Intel a PR nightmare and a $500m charge.
I remember the bug ... but I don't remember me no Pentium 5. Man that caused heartache for us (I was working at a Uni supporting people doing mathematical modelling - we were directly affected). But it was a Pentium (meaning 5th) which followed the 286, 386 and 486 chips. It would have been the 586 but Intel had been told they couldn't protect the numbers as names, and thus we had had the AMD 386, 486 etc. directly competing with the Intel parts.
And Intel just couldn't take it :)
- Mounties get their man: Heartbleed hacker suspect, 19, CUFFED
- Batten down the hatches, Ubuntu 14.04 LTS due in TWO DAYS
- Samsung Galaxy S5 fingerprint scanner hacked in just 4 DAYS
- Feast your PUNY eyes on highest resolution phone display EVER
- Wall St's DROOLING as Twitter GULPS DOWN analytics firm Gnip