Well this is awkward. As Microsoft was bragging about Office at Build, Office 365 went down Microsoft scrambled to get its Office 365 service back online after the cloud productivity suite was hit with a mid-day outage. Loads of people reported that, at around 1245 PT, access to the service went out. Microsoft confirmed shortly after it was having problems, and said it was looking into the matter. Subscribers in New …
But here's the sad irony. As more and more people move to this model, when one subscriber is down, they're all down. No finger pointing about how cheap/dumb/careless you are to use O365 (or any other cloud provider) and how the in-house service you use is still up, because the guy that would point that out is down too. So for MS/Amazon/Salesforce etc, the more people in the cloud the more immunity from criticism their subscribers get. I predict a no-work-gets-done-anywhere day in the not-too-distant future, and everyone will just shrug. Someone start sharpening the pencils.
Given the pro on prem kit these comments are showing, I am expecting massive downvotes.
Having worked with Cloud products for near 10 years now, I can say that in general they are much more reliable and lower cost than an on prem equivalent. Sure you cannot stroke the tin, though you do get a much better infrastructure from a company that invests millions/billions and has SLA's around it all than the bit of tin you got from the reseller round the corner.
"SLA's around it all"
Do those SLAs actually prevent things from going down?
Do they actually provide compensation for the real costs to the client when they do go down?
Do those trying to fix things when they go down have your, the client's interests as their prime motivation or are they just working to the SLA?
Cloud is more swings and roundabouts; it shouldn't be your belt and braces. What are the advantages, disadvantages and rationales behind using Cloud for your business? Is it appropriate? If it's an advantage, appropriate, apposite and not business critical, it's a no brainer, go for it, it's not going to kill you, and it will probably help you. If it's critical to your business, not apposite, not appropriate and confers no advantage, do it on-prem - that's also a no-brainer.
The real sticking point is when the advantages are small but present, it could be achieved either on-prem OR in-cloud with relatively little difference in cost, and where there's ambiguity about appropriateness. The key deciding factor then is how critical is that function to your business?
"Having worked with Cloud products for near 10 years now, I can say that in general they are much more reliable and lower cost than an on prem equivalent".
Perhaps in general but NOT for us. Since moving to Office365 last year (not my idea) we have had multiple DAILY problems. Users are requested to sign in over and over, random "the network cannot be accessed, try again later", endless connection issues like this. Simple things don't work, like clicking a PDF attachment and it opens with your PDF viewer. When the wheels fall off, we can do nothing but wait for Microsoft to figure it out. It's a crap experience for my users, who don't trust it.
Previously we ran it all from here and it worked brilliantly, daily, for years. When it broke, we fixed it quick and back in business. Properly maintained, on-premises kicks the crap out of clown, sorry, cloud offerings.
We moved to it 3 months ago after an 18 month trial with 50 users and we've now got 500 users on it out of 900 and they prefer it. The helpdesk prefer it as they're dealing with less local install issues or MS apps. The purchase dept like it 'cos the licensing issues are a lot simpler to deal with. The IT techs love it 'cos they can use an O/S they like, they're not beholden to an Office install, we have dekstops of Linux, Solaris, OSX and Windows and they can all use the browser based Office apps and share docs. The biggest advantage to users is that they haven't got faff about with VPNs and RSA keys to get to their mails or docs while out of the office, they simply jump on a secure app on their phones or laptops, or they use a public browser to get emails quickly. The management love 'cos it they can bother the staff, the staff love it 'cos they know their managers have no excuses not to answer emails and get decisions made. Alright once in a blue moon it goes offline but cloud services are still in the learning stage, and I would say that cloud is not for everyone but basic stuff like word docs, spreadsheets and emails worked out well for us.
"Having worked with Cloud products for near 10 years now, I can say that in general they are much more reliable and lower cost than an on prem equivalent. "
OK so far, though not exactly detailed.
""I am expecting massive downvotes.""
OK. Now, why might that be?
Normally I'd be courteous and ask whether you might have an interest to declare. But you've helped us out so much already that the question doesn't even need asking, thank you:
"I am on the GFI sales team though within the MSP "MAX" product range based in Dundee, cloud is really well received though if you are looking for something to manage clients as a support company instead of end users within your organisation you want GFIMAX Remote management, Cloud was built to be the "Sys admin" MAXRM."
https://forums.theregister.co.uk/forum/1/2014/05/06/gfi_cloud_review/
Let the downvotes continue, now for twice as many reasons.
"Having worked with Cloud products for near 10 years now,"
You're a cloud salesman. You're misleading El Reg readers, now or in your previous posts - take your pick. Downvotes deserved.
"you do get a much better infrastructure from a company that invests millions/billions and has SLA's around it all than the bit of tin you got from the reseller round the corner."
As others have noted here already, SLAs are worth their weight in gold. IE worthless. Competent delivery is worth more than an SLA but competent delivery doesn't come easy, doesn't come free with an SLA, and isn't necesarily immediately available by just outsourcing the business to some cheapskate outsourcer which now calls itself a cloud outfit. And by definition, cloud outfits are cheapskate - cheap is what they're selling.
As others will have noticed round here already, MS UK Azure sales opportunities are currerntly being referred to MS Canada capacity, right? Won't be the first, won't be the last, won't just be MS.
https://www.theregister.co.uk/2017/05/04/microsoft_azure_capacity_woes_hit_uk_customers/
On the other hand, how long does it take to get a small vanload of servers and storage from a random Dell/HP/whatever boxshifter?
Wow, so all sales people are bad right?
Yes I work in sales, though I trained in IT/Networking and that is where I started, and while in Sales I am of the technical sales, so giving demos and the like.
Hate sales all you want, but without sales people then nothing would progress and most people would not have jobs, also the best sales people actually tell the truth, especially in a SaaS/Cloud model since if you sell on a lie people can cancel right away so it would only bite you in the arse.
Now does physical kit have a place and have pros? Yes. Does that mean Cloud is pointless and bad? No, Somethings are just better suited for the cloud and having physical systems for them is counter intuitive. Tech progresses, learn to move with it, or be left in the past.
Also notice how I don't hide behind Anonymous coward?
I thought the number was referring to the number of directions that the applications could be approached from, you know, like the current trend for performance feedback. It provides 365 degree coverage, i.e. a complete circle (did they ever get that floating point bug fixed?)
This is why effectively centralised, subscription (rentier), internet resources are stupid-fragile slavery, and people should not to use them for critical stuff where downtime can't be worked around using local resources; much like centralised (fragile) SVN is inferior to local and distributed (anti-fragile) git!
The icon because the children may later have to deal with more of this rentier fragility and slavery, and may have less local fall-back resources, because of this subscription-business-model trend by corporations.
Remind me again why I should trust a company with centralized control of my data ... Especially when that company spent decades trying to move control of the personal desktop from mainframe data centers to the personal computer?
No, thank you. I'll keep it in-house. For values of "in-house" that include a couple continents. Honestly, it's not all that hard to roll your own.
"Remind me again why I should trust a company with centralized control of my data"
You should trust them because the penalty clauses in the contract make it really bad for them if you suffer any kind of outage and so they'll make every conceivable effort to deliver. Just like any other kind of service or product that you buy in from outside, in other words. Businesses have been doing this for years and I really don't see what the fuss is about.
Obviously it would be bonkers if you didn't have any such clauses in the contract ...
You should trust them because the penalty clauses in the contract make it really bad for them if you suffer any kind of outage and so they'll make every conceivable effort to deliver.
Trouble is enforcement and getting timely recompense, ie. before your business (or their business) collapses. The current HBOS fraud case gives a good idea of the time lines: 2002~2007 fraud against small business customers committed, concerns raised, Police investigation started 2010, trial of 6 people commences 2016 with convictions Jan 2017, it only now that compensation for the victims is being discussed, yet it is likely these will also involve court proceedings so it could be a few more years before victims receive any monies... Also bear in mind in the aftermath of the 2008 financial crash HBOS itself failed and only survived by Lloyds taking on all its business and liabilities...
yah, they have penalty clauses.
it's like knowing your next of kin can sue after you're dead.
Doesn't mean d*ck to me.
here, we spent million$ to go from an Exchange system central to each Agency to a O365 setup for all Agencies. But it's actually "local" as it all physically lives central to most Agency locations. So basically it's Exchange centralized.
So pretty much what we had the capability of BEFORE but it's new, shiny and "cloudy" for millions of dollars invested.
We shoulda called our on-site server room a "cloud" as it was just as "cloudy" as our current O365 setup.
This post has been deleted by its author
sadly it happens so much it's like making fun of special needs kids who have balance or coordination issues. the failures of one product or the other are almost ongoing.
the humor comes from having stories like these to forward to the O365 Evangelicals who claimed that this sort of thing would happen LESS than on a "traditional" locally administered system.
The excuses they generate are where the humor lies.
Universities switched to Office365 because its basically free. Of course it doesn't offer the same kind of performance the old on-prem exchange, but running a resilient mail system for 30,000 students isn't cheap, backing it all up is also very expensive. You can probably save £1m over 5 years by making the move.
it won't be "free" in 5 years. Why TF do you think a mega corp would give you free stuff indefinitely?
They're all racing to the bottom until they reach critical mass, then watch the prices rocket. And there will be few indy hosting companies left, so what ya gonna do then? 'Cos Ghostbusters are no use.
Because the idea is that they hook the students on office 365 so that when they leave school or uni they'll push office 365 at the business they start or go to work for.
It's the same reason we get super cheap volume licences from Microsoft.
The alternative is our network goes Apple or Linux and we have thousands of converts leave the institution each year who won't want to use a Microsoft product.
"... but that would mean management admitting they made a mistake."
Bingo. Some of ours will swallow all of the marketing bilge they're given, and a lot of that is second-hand evangelising by our own managers who have already swallowed it themselves.
When things go wrong or performance suffers and staff start to complain, management explain the service is cheaper... constantly updated... new features all the time... etc, whereas had the same issues occurred when on-prem, those same managers would have been pointing the fingers and demanding action as IT were not providing the service the should.
@Ken - "Perhaps they felt that providing a map, with state boundaries and fuzziness in affected areas..."
I'm in one of the two fuzziness areas, as are the vast majority of Texans. The scare headline "Office 365 went down - TITSUP: Total Inability To Stand Up Products" is a bit overdone if you ask me. I'm sure it was good for the clicks though.
nobody disputes that the area covered by Texas is larger (nearly 3 times larger) than the area covered by the UK. But it kind of overlooks the fact that has less than half the population of the UK.
Microsoft use distributed locations to host their authentication servers - so different people in different places will authenticate to different servers - is that too much to grasp?
Obviously this was not a global outage - it was an issue with some of the authentication servers in some locations - thereby affecting some users.
A vanilla generic cloud offering based out there somewhere is unlikely to match a well run local service.
On the other hand it is unlikely to go as spectacularly bad as a dreadfully run local service (looking at you UCL).
In time the level of service will be accepted and IT skills will wither on the vine.
Then prices will start to rise.
In time the level of service will be accepted and IT skills will wither on the vine.
Then prices will start to rise.
AKA "The Walmart Strategy[1]"
[1] - Yes - the one they used extensively in the US - move into an area where they were not really needed, use their financial muscle to drive local shops into bankruptcy and then close that loss-making store so that locals now have drive to the nearest supermarket. Which will be, surprise, surprise, a Walmart..
"User Impact: Users may have been unable to log in to Office 365 services.
Final status: We monitored the service and worked with some affected customers to confirm restoration. In some cases, affected users may need to restart their browser or application in order for sign-in to begin functioning properly.
Scope of impact: Customer reports indicated that many users likely experienced impact related to this event. Our analysis indicated that this issue could have potentially affected any of your users intermittently if they were routed through the affected infrastructure.
Start time: Wednesday, May 10, 2017, at 7:32 PM UTC
End time: Wednesday, May 10, 2017, at 8:12 PM UTC
Preliminary root cause: A recent change introduced a configuration issue that caused authentication for Office 365 services to become degraded.
Next steps:
- We're reviewing our monitoring services to find ways to reduce detection time and to improve our automated recovery processes.
- We're reviewing our update procedures to help catch this kind of problem during our testing cycle.
We'll publish a post-incident report within five business days."
So someone made a change and it went wrong. It happens.
Sysadmins who manage on-prem, private clouds are an aging population. they are fed up with the constant refresh cycle, fw updates, failures, backups, testing, security exploits, change control and the countless other waist of time activities that are invisible to the business and add no real value. Bunging all that backend work into the cloud for some other spod to manage isn't such a bad idea. They can then get on with developing better services that actually add value (and not have to respond to that dreaded alert saying the SAN is down). But don't get me wrong, all eggs is a valid argument, your critical business systems should be hybrid or hosted on 2 different public clouds, just like you'd design fault domains into your on-prem infrastructure.
If email is critical to your organisation, then don't outsource it to a public cloud provider. I don't really see why everyone is laughing/attacking the service. Its email, not life support.
If your business relies so heavily on email that potential uncontrollable downtime from a public cloud provider, don't migrate. Invest in your own system and sysadmins.
To the rest of us, where email being down doesn't mean the end of the world, we're quite happy with a public cloud provider.
so here's my primary "beef":
we are a public service Agency that has been in operation for nearly a century.
None of the companies offering "cloud based services" have even been in existence for half that, and none have offered or maintained any of these services for a tenth of that.
Very few companies, if any, have survived that long and maintained the same business services. the IBM of today is not the IBM of its inception, nor the IBM of even a decade ago.
So sure we need to adapt to new technologies for efficiency reasons, but making fundamental core business processes 100% dependent on exterior private corporations with less than 5 years actual deployment of those services? Or in some cases companies that are straight out of the Startup Incubator? Where if those services become unavailable the time and resources needed to restore or transfer functionality to some other provider are WEEKS and $million$ in which vulnerable populations are completely screwed?
At least with internal email, when things stopped, we could find admins to blame. When the "cloud service" company goes out of business, disappears and its offices are empty, we can lay blame but we can't restart their servers and provide services to clients.
[terminally insufficient throughput segregates usable product]
It's sad how people are flocking backwards to before the earliest PC days. Seems like you never know when a golden age *is*, only when it *was*. Devo must be having a good laugh, though.
(it's that one with a Zaurus in the pocket)
For sure not anybody younger than Generation Z,
Most webmail clients do a decent job for formatting nice letters. When it has to be fancy, wiki text editing and pasted into an email does the job.
For the rest, all this MS stuff is as modern and relevant as a Remington typewriter from the 1930's, even when it has 500 buttons more.
Bandwidth challenges in my workplace mean four seconds from selecting a mailing to it actually loading into the preview pane when everything is working properly. Periodically, typically as I'm trying to type a reply or building a filter to move something important out of the storm of reply-to-all "Me too"s which liven up my day, the spinning wheel of annoyance pops up and the entire shambles goes Nonresponding Grey for anything up to 15 seconds.
This outage just meant I was less aggravated by the bloody thing yesterday.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
