App Store, Apple Music, iCloud & other Apple services impacted by Amazon outage [u]

Posted:
in iCloud edited February 2017
A variety of Apple cloud services experienced outages and slowdowns on Tuesday in what appeared to be a major issue with Amazon Web Services, which hosts some of the biggest online companies in the U.S.




The issue is affecting some users of Apple Music, the Apple TV, and the Mac and iOS App Stores, plus a number of iCloud services including Notes, backup, web apps, and iCloud Drive, according to Apple's system status page. People may also be seeing trouble with iTunes, iBooks, and/or cloud-based Photos functions.

Problems began shortly before 1 p.m. Eastern time and are still ongoing.

While Apple hasn't acknowledged the source of the glitches, Amazon's S3 (Simple Storage Service) is currently experiencing "availibility issues" on the U.S. east coast, leading to problems for dependent companies. Apple is known to rely on Amazon for portions of its cloud infrastructure, despite investing in large-scale data centers in places like North Carolina.

The company is thought to be gradually shifting away from its dependence on Amazon. It will, for instance, be turning its Mesa facility into a "global command center," and is working to establish its first European data centers in Ireland and Denmark. The Irish project has yet to begin construction, owing to delays caused by local concerns like environmental impact.

Recently it was revealed that Apple is planning to expand its Reno data center by over 375,000 square feet, at a cost of about $50.7 million. The facility currently handles tasks like Siri, FaceTime, and iMessage.

Update: Apple's system status page is showing an all clear as of 1:35 p.m. Pacific.
«1

Comments

  • Reply 1 of 25
    There's been a definite sign of problems as far as updating apps. It's slower than molasses in January.
    napoleon_phoneapartanton zuykovsergioz
  • Reply 2 of 25
    Why doesn't Apple host all its own business on its own servers!
    napoleon_phoneapartanton zuykov
  • Reply 3 of 25
    Ironically it seems to have gotten down detector as well. 
  • Reply 4 of 25
    Apple Music and iTunes Match have not been working properly since about 10:00isn PST this morning. While Apple will not publicly discuss, their Internet Services Operation (ISO) is a user of Amazon Web Services and thus Apple users have likely been affected by this issue.

  • Reply 5 of 25
    Why doesn't Apple host all its own business on its own servers!
    They have been moving rapidly to move most if not all of their content to 'in-house' servers. Building and maintaining a huge Cloud infrastructure is no trivial task and utilizing outside services, such as Amazon, Microsoft, and Google, provides them time-to-market and flexibility that they might not otherwise have by strictly relying on their own. I work in this business and understand some of the challenges and complexities of building and maintaining a highly reliable, scalable, and flexible infrastructure. Apple is good at many things, but this has NEVER been a particular strength of theirs...
  • Reply 6 of 25
    sog35 said:
    This right here is proof that serious companies (including Apple) should have multiple Cloud providers. 

    If Amazon is down? Well you have a backup.

    Hope more corps realize this and start diversifying away from Amazon cloud
    sog35 said:
    karmadave said:
    Why doesn't Apple host all its own business on its own servers!
    They have been moving rapidly to move most if not all of their content to 'in-house' servers. Building and maintaining a huge Cloud infrastructure is no trivial task and utilizing outside services, such as Amazon, Microsoft, and Google, provides them time-to-market and flexibility that they might not otherwise have by strictly relying on their own. I work in this business and understand some of the challenges and complexities of building and maintaining a highly reliable, scalable, and flexible infrastructure. Apple is good at many things, but this has NEVER been a particular strength of theirs...
    Why is it so hard to build a cloud?

    I mean if Apple throws Billions into cloud, would they not have a cloud just as good as Amazon? 
    You really should avoid commenting on things you know absolutely nothing about. Just my opinion.
    SpamSandwichtmayyoyo2222sergiozelijahgsingularitytdknoxRayz2016chiawatto_cobra
  • Reply 7 of 25
    Amazon had better not have too many big issues with their web services division. The profits from AWS are subsidizing a lot of their unprofitable business lines, and without a healthy AWS it's going to be very hard (as it already should be) to justify a P/E that is more than 10x Apple's.
    watto_cobra
  • Reply 8 of 25
    volcanvolcan Posts: 1,799member
    Why doesn't Apple host all its own business on its own servers!
    They don't have the infrastructure yet. It takes a long time to scale up especially since they want to do it with renewable energy. For every data center they build, they need to build the power infrastructure as well.

    I don't think they particularly like AWS because it is kind of unreliable and slow, but they do have 42 data centers around the world. Microsoft has more than 100. They also use Akamai which has 50 data centers. Until Apple can get to a similar scale they have no choice but to use those third party providers. Apple has nearly a billion users and no where near enough infrastructure to serve them with only 4-5 data centers that they own, all in the US. Their China data centers are owned by the Chinese. Not sure how many there are there. Apple is planning to build 2 new data centers in Europe.


    wonkothesane
  • Reply 9 of 25
    lkrupplkrupp Posts: 10,557member
    So bottom line according to some here, even though it’s Amazon’s servers it’s still Apple’s fault. It’s aways Apple’s fault isn’t it. 
    lolliverwatto_cobra
  • Reply 10 of 25
    sflocalsflocal Posts: 6,092member
    So Sog is now an sofa-taught Systems Engineer in a position to tell Apple how to build redundant, scalable data centers?  Wow... he's a child wearing many fuzzy animal hats.
    edited February 2017 Rayz2016lolliver
  • Reply 11 of 25
    sog35 said:
    This right here is proof that serious companies (including Apple) should have multiple Cloud providers. 

    If Amazon is down? Well you have a backup.

    Hope more corps realize this and start diversifying away from Amazon cloud
    Apple DOES use multiple cloud providers, however a failure at one doesn't necessarily mean they can just failover to another. Multiple cloud providers doesn't provide failover/redundancy. It provides capacity...
    sog35 said:

    Why is it so hard to build a cloud?

    I mean if Apple throws Billions into cloud, would they not have a cloud just as good as Amazon? 
    I don't even know where to begin. Unless you have worked for, or sold to, a Cloud provider it's difficult to understand the complexity and enormity of the task. While Apple does have plenty of $$$ to throw at the problem, recruiting and maintaining the talent is equally (if not more) important. In case you haven't noticed, two of the three major Cloud service providers are based in Seattle. Not the SF Bay Area. Finally, it takes time to find locations, build the necessary datacenter capacity, procure and standup infrastructure, etc.

    Apple's expertise is in PLATFORM and not Infrastructure. 

    lolliver
  • Reply 12 of 25
    volcanvolcan Posts: 1,799member
    karmadave said:
    Apple DOES use multiple cloud providers, however a failure at one doesn't necessarily mean they can just failover to another. Multiple cloud providers doesn't provide failover/redundancy. It provides capacity...
    Multiple data centers can provide backup. Many companies do this, like Facebook for example. Of course they own their own redundant data centers, but it could be done even across different providers. The difficulty in Apple's case is that MS provides Azure database services and Amazon some other types of content, so they cannot really sync because they are essentially different platforms, but you could theoretically sync between two MS data centers or two Amazon centers. With a catastrophic failure it takes quite some time to have the backup data center take over the master data center duties. It doesn't happen immediately because of TTL DNS. Usually by the time it does take over, the master data center can be brought back online.
  • Reply 13 of 25
    lkrupplkrupp Posts: 10,557member
    Why doesn't Apple host all its own business on its own servers!
    Well, people bitch and whine every time there’s a hiccup with any Apple service. Out come the memes about Steve being dead, Apple quality on the decline, you know, the usual mindless, whining pablum. So pick your poison. If it’s Amazon’s servers, it’s Apple fault. If it’s Apple servers, it’s Apple’s fault. If the weather changes, it’s Apple’s fault.
    watto_cobra
  • Reply 14 of 25
    BTW, my company was also affected by this outage which was apparently caused by a Global Denial of Service (GDOS) attack...
  • Reply 15 of 25
    volcan said:
    karmadave said:
    Apple DOES use multiple cloud providers, however a failure at one doesn't necessarily mean they can just failover to another. Multiple cloud providers doesn't provide failover/redundancy. It provides capacity...
    Multiple data centers can provide backup. Many companies do this, like Facebook for example. Of course they own their own redundant data centers, but it could be done even across different providers. The difficulty in Apple's case is that MS provides Azure database services and Amazon some other types of content, so they cannot really sync because they are essentially different platforms, but you could theoretically sync between two MS data centers or two Amazon centers. With a catastrophic failure it takes quite some time to have the backup data center take over the master data center duties. It doesn't happen immediately because of TTL DNS. Usually by the time it does take over, the master data center can be brought back online.
    Volcan. You totally get it. As I said above, this was a Global Denial of Service (GDOS) attack and it looks it's mostly been resolved. Thanks! 
  • Reply 16 of 25
    asdasdasdasd Posts: 5,686member
    sog35 said:
    karmadave said:
    sog35 said:
    This right here is proof that serious companies (including Apple) should have multiple Cloud providers. 

    If Amazon is down? Well you have a backup.

    Hope more corps realize this and start diversifying away from Amazon cloud
    Apple DOES use multiple cloud providers, however a failure at one doesn't necessarily mean they can just failover to another. Multiple cloud providers doesn't provide failover/redundancy. It provides capacity...
    sog35 said:

    Why is it so hard to build a cloud?

    I mean if Apple throws Billions into cloud, would they not have a cloud just as good as Amazon? 
    I don't even know where to begin. Unless you have worked for, or sold to, a Cloud provider it's difficult to understand the complexity and enormity of the task. While Apple does have plenty of $$$ to throw at the problem, recruiting and maintaining the talent is equally (if not more) important. In case you haven't noticed, two of the three major Cloud service providers are based in Seattle. Not the SF Bay Area. Finally, it takes time to find locations, build the necessary datacenter capacity, procure and standup infrastructure, etc.

    Apple's expertise is in PLATFORM and not Infrastructure. 

    offer those same employees a 200% raise and they dump Amazon/MSFT in a heart beat.

    Again with $$ you can do almost anything.
    Even of you got all engineers you still need to scale up all your data centres. Then write the middle ware and the  database and redundancy handling that Amazon have ready built and own the ip for (and since you have the employees they will win if you copy anything). 

    And you won't get all the engineers. 

    There's nothing harder than getting into cloud. 
  • Reply 17 of 25
    asdasdasdasd Posts: 5,686member
    lkrupp said:
    Why doesn't Apple host all its own business on its own servers!
    Well, people bitch and whine every time there’s a hiccup with any Apple service. Out come the memes about Steve being dead, Apple quality on the decline, you know, the usual mindless, whining pablum. So pick your poison. If it’s Amazon’s servers, it’s Apple fault. If it’s Apple servers, it’s Apple’s fault. If the weather changes, it’s Apple’s fault.
    Nobody blamed apple. You came in defending them. 
  • Reply 18 of 25
    lkrupp said:
    So bottom line according to some here, even though it’s Amazon’s servers it’s still Apple’s fault. It’s aways Apple’s fault isn’t it. 
    Actually, these days it is usually the Russians' fault. 
    bestkeptsecret
  • Reply 19 of 25
    volcanvolcan Posts: 1,799member
    karmadave said:
    Volcan. You totally get it. As I said above, this was a Global Denial of Service (GDOS) attack and it looks it's mostly been resolved. Thanks! 
    Yep! All Apple services are back up.

     Apple's system status page

    DDOS attacks usually cannot be resolved by the engineers at the data center. They often need to call in specialized Cisco engineers who can analyze the packet signature of the attack and write a specific filter, then install it on the edge routers. That takes time.
    edited February 2017
  • Reply 20 of 25
    The Russians are coming!  The Russians are coming!  Oh, wait...
Sign In or Register to comment.