[HN Gopher] Operational issue - Multiple services (UAE)
       ___________________________________________________________________
        
       Operational issue - Multiple services (UAE)
        
       Author : earthboundkid
       Score  : 142 points
       Date   : 2026-03-01 19:24 UTC (3 hours ago)
        
 (HTM) web link (health.aws.amazon.com)
 (TXT) w3m dump (health.aws.amazon.com)
        
       | potatoproduct wrote:
       | We are living in increasingly weirder times.
        
         | dgxyz wrote:
         | Not really. It's just been pretty damn quiet for years.
        
         | debo_ wrote:
         | When I first learned that there were AWS Middle East regions,
         | my first thought was "wow they are more optimistic than I am ."
        
           | Cyph0n wrote:
           | No, they are more aware of the customer demand for compute in
           | the region.
        
             | alexfoo wrote:
             | And demand for data sovereignty.
        
               | Cyph0n wrote:
               | Absolutely, especially in the KSA.
        
           | toast0 wrote:
           | Google Cloud also has middle east locations. As does Azure,
           | Oracle and Alibaba. Afaik, IBM Cloud does not. I think those
           | five and AWS are the top 6 global public access clouds.
        
         | astrange wrote:
         | A factory not working because of a missile strike seems pretty
         | classic actually.
        
           | guerrilla wrote:
           | Sure, but it's not a factory.
        
             | mediaman wrote:
             | Why not? It's a physical building with lots of equipment
             | that produces products shipped to its customers.
             | 
             | Its products are sequences of electrons, instead of atoms.
             | But so are power plants. And in the context of what happens
             | when they're hit by missiles, a factory, data center, and
             | power plant all behave the same.
        
             | astrange wrote:
             | It's a big building with a lot of capital assets inside
             | that are the means of production for a business...
        
       | Shank wrote:
       | I wonder if this was bad targeting job or intentional. I
       | appreciate the transparency and optimism in the status updates
       | though!
        
         | sb057 wrote:
         | Looking at Google Maps, there's Al Dhafra Air Base a couple of
         | miles to the datacenter's south, an oil refinery a bit to the
         | east, ports to the north, and a military academy to the west.
        
       | general1465 wrote:
       | In Southern Europe some smaller web servers are intermittently
       | not working, while big servers like YouTube are working fine. But
       | I don't think it is related.
        
       | Trasmatta wrote:
       | Is this the one in Bahrain?
        
         | the_mitsuhiko wrote:
         | UAE. Abu Dhabi
        
       | ejdyksen wrote:
       | Just one AZ, not the whole region:
       | 
       | > The other AZs in the region are functioning normally. Customers
       | who were running their applications redundantly across the AZs
       | are not impacted by this event.
        
       | boxedemp wrote:
       | Amazon usually has 3 AVs per region, looks like there are
       | surviving AVs but the system didn't switch over gracefully.
       | 
       | I bet that was an interesting sev2 ticket!
        
         | easton wrote:
         | It depends on the service if things move gracefully or not. The
         | incident explains it's only EC2 (and dependent services) in
         | that AZ, so if they try to route traffic for services hosted on
         | EC2 to that AZ it's not working (and customers running
         | instances in that AZ have lost access).
         | 
         | The other ones are not impacted. They always like to tell you
         | to pay for more than one instance in different AZs so if this
         | happens you don't get impacted.
        
       | eptcyka wrote:
       | > one of our Availability Zones (mec1-az2) was impacted by
       | objects that struck the data center, creating sparks and fire.
       | 
       | God forbid we'd ever say that it was struck by a missile or a
       | munition in an act of war.
        
         | NikolaNovak wrote:
         | That's what I'm trying to understand too. It's this a
         | meteor,tree,etc? Or a human made object,and if so accidental or
         | intentional one. Further risk assessment would be dependent on
         | root cause.
        
           | rrvidqdi wrote:
           | The earliest phrasing I saw internally was "Root cause is
           | identified as a drone attack to DXB61 site". That's somewhat
           | open to interpretation, and could also have simply been
           | incorrect. It was scrubbed from the ticket, though, and it
           | now merely vaguely gestures toward a "power event". The
           | ticket I'd expect to have further detail was locked down.
        
         | bigyabai wrote:
         | Potential moon bear attack; we're waiting on satellite imagery
         | to confirm it. https://youtu.be/pvjgIxuVdo4?t=96
        
         | hdgvhicv wrote:
         | Maybe a missile, maybe a drone, maybe debris
         | 
         | Doesn't really matter, we know trumps latest war is the cause
        
         | arjie wrote:
         | I actually like the way they said it. I don't know if it's a
         | different cultural tradition, but the cool steely-eyed fact-
         | based conversation always really felt so much more inspiring:
         | Conrad: I got three fuel cell lights, an AC bus light, a fuel
         | cell disconnect, AC bus overload 1 and 2, Main Bus A and B out.
         | Aaron: Flight, EECOM. Try SCE to Aux.
         | 
         | Modern culture in the movies and whatnot is that someone should
         | be yelling "Everything's failing. Give me something, Houston.
         | All lights are on! MAYDAY MAYDAY!" and some sort of flavour
         | commentary like that. But reading engineering updates that go
         | like this feels like watching maximal professionalism under
         | fire:
         | 
         | > _At around 4:30 AM PST, one of our Availability Zones
         | (mec1-az2) was impacted by objects that struck the data center,
         | creating sparks and fire. The fire department shut off power to
         | the facility and generators as they worked to put out the fire.
         | We are still awaiting permission to turn the power back on, and
         | once we have, we will ensure we restore power and connectivity
         | safely. It will take several hours to restore connectivity to
         | the impacted AZ. The other AZs in the region are functioning
         | normally. Customers who were running their applications
         | redundantly across the AZs are not impacted by this event. EC2
         | Instance launches will continue to be impaired in the impacted
         | AZ. We recommend that customers continue to retry any failed
         | API requests. If immediate recovery of an affected resource
         | (EC2 Instance, EBS Volume, RDS DB Instance, etc.) is required,
         | we recommend restoring from your most recent backup, by
         | launching replacement resources in one of the unaffected zones,
         | or an alternate AWS Region. We will provide an update by 12:30
         | PM PST, or sooner if we have additional information to share._
         | 
         | This has that same mechanical tone of an ice-cold captain
         | dealing with a proximate situation providing exactly the
         | information they know. No flavour commentary. Amazing. I
         | fucking love it.
        
           | jcgrillo wrote:
           | High signal/noise ratio is extra important when things are
           | going badly
        
           | AlexErrant wrote:
           | For those who wanna hear it
           | https://youtu.be/-rqL035klC4?t=101
           | 
           | (Lightning at 1:03)
        
       | p-o wrote:
       | Interesting adjacent theory is how much are datacenters becoming
       | military target to strike as part of disrupting initial defenses.
       | It doesn't seem it was the case in this instance, but I could see
       | this becoming a more important target in future.
       | 
       | Seems like it should be somewhat easier to bomb 50 datacenters
       | than it would be to hack and disrupt 1000s of different services.
       | 
       | Again, this is just me thinking out loud on a tangent and this
       | doesn't have much to do with this story, but I felt it was an
       | interesting thought to share nonetheless.
        
         | Imustaskforhelp wrote:
         | > Seems like it should be somewhat easier to nuke 50
         | datacenters than it would be to hack and disrupt 1000s of
         | different services.
         | 
         | The bigger part of me seems that if we someone nukes 50
         | datacenters all at once or say all of Amazon's datacenters at
         | once, then the data stored in there would simply be gone and
         | given so many datacenters are located in Virginia,USA iirc or
         | just so many companies being reliant on few datacenter
         | providers.
         | 
         | The larger threat to me with the lose of data is firstly the
         | panic within public fronting services but also, with Hedge
         | Funds, Pension funds or banking datacenters who might be using
         | these and if they lose the data, then its gonna cause even more
         | public mayhem.
         | 
         | Some might be saying oh off-site backups exist but there has
         | atleast been one instance, where a single Google accident had
         | led to massive issues for a 135 Billion $ pension fund.
         | 
         | Relevant Kevin Faang video about it:
         | https://www.youtube.com/watch?v=3GOAUyipnM4 [Google
         | Accidentally Deletes $135 Billion Pension Fund, Chaos Ensues]
        
         | tbrownaw wrote:
         | > _Seems like it should be somewhat easier to nuke 50
         | datacenters than it would be to hack and disrupt 1000s of
         | different services._
         | 
         | Previous outage news makes it sound like the cloud providers
         | still have quite a few logical single points of failure.
        
         | Zeyka wrote:
         | That's so interesting. Are any of the US military (or other
         | satellite state of the US) systems running in "normal"
         | datacenters or do they have a few protected DoD datacenters in
         | the US?
        
           | Imustaskforhelp wrote:
           | Found this relevant article:
           | https://serverlift.com/blog/military-modular-data-centers/
           | (AWS Military Modular Data centers)
           | 
           | I do think that though, atleast from the Anthropic decision
           | prior, we know that Anthropic which was used by DoD should be
           | on normal AWS datacenters.
           | 
           | I am saying this because, Dod Threatened to force take the
           | source code of Anthropic if they don't agree to aggregious
           | demands so that means that they don't have the source code.
           | 
           | Perhaps DoD used Anthropic within AWS Military modular DC's
           | but I find it extremely unlikely.
           | 
           | I am almost certain that even with OpenAI who bent its knee
           | to DoD, its still hosted on regular infrastructure and DoD is
           | using these AI models on pretty sensitive tasks (During the
           | Venezeula Maduro's capture, Anthropic/Claude were used iirc
           | to handle some data analysis)
           | 
           | IMO Tho, Any Employee from Anthropic/OpenAI might know better
           | tho about how these models are deployed.
        
             | riknos314 wrote:
             | they have a press release stating that Claude is available
             | on bedrock in secret regions
             | 
             | https://aws.amazon.com/about-aws/whats-new/2025/11/amazon-
             | be...
             | 
             | https://aws.amazon.com/federal/secret-cloud/
        
         | jcgrillo wrote:
         | IIUC part of the reason ballistic missiles have multiple
         | warheads is that some of them detonate high up to knock out air
         | defenses and other electronics allowing the rest to fall
         | through to their targets. The last time we tried this
         | experiment as a species was the starfish prime tests in 1962
         | which caused some electrical havoc in Hawaii. These days our
         | systems are probably more delicate and sensitive? All that is
         | to say, in a scenario where nukes are going off I'm not sure
         | you'd even need to target any datacenters in particular..
         | they're probably all toast by default.
        
         | swiftcoder wrote:
         | The more interesting question, is how many datacenters are just
         | plonked next to a high-value military target?
         | 
         | For infrastructure reasons, we plonk datacenters down next to
         | airports big enough to fly major hardware into, and near where
         | the big oceanic cables come ashore... and for strategic reasons
         | those are also the perfect places to place military bases
        
           | throw475787 wrote:
           | Is there acrually some meaningful physical separation between
           | military and civilian server deployments?
           | 
           | We seem to be really bad at separating those two. For example
           | Starlink is basically military infrastructure now, used to
           | guide bombs.
        
           | cherryteastain wrote:
           | A datacenter IS a high value military target.
        
         | roxolotl wrote:
         | This is the data center version of https://xkcd.com/538/.
         | Realistically if there is a hot war what you're saying seems
         | accurate.
        
         | roncesvalles wrote:
         | Exactly. 2 is only sufficient for HA against random failures.
         | It's not enough for HA against a determined adversary willing
         | to use targeted force.
        
       | anonu wrote:
       | We have business in UAE. For whatever reason I defaulted to us-
       | west-2 since these particular applications are not latency
       | sensitive.
        
       | Imustaskforhelp wrote:
       | Has this ever happened ever in history of Cloud providers before
       | this because of war?
       | 
       | They mention that the datacenter had fires and sparks and they
       | are mentioning hours of downtime but given the situation, How
       | does that prevent the situation from happening again. It's best
       | for people to use safer regions than the middle east in the
       | moment as missiles might target the same datacenter seeing that
       | some damage was caused.
       | 
       | Moving forward, will there be a demand (all be small) for nuclear
       | bunker esque datacenters which can withstands missiles? I know
       | absolutely nothing about constructing underground but can
       | explosives not be used to create underground datacenters
       | comparatively cheaply? One can also use revamped Nuclear bunkers
       | (although the scale of AWS datacenters might be huge tho who
       | knows)
       | 
       | Had some ideas which show that this idea might be interesting,
       | https://www.nature.com/articles/s44284-026-00406-2
       | 
       | I am curious but what are the safety attempts made by Internet
       | Exchange Providers or (had to search it up) but Submarine Cable
       | landing stations, to me it feels like blowing these up leads to
       | internet downtime across whole country / between providers.
        
         | tbrownaw wrote:
         | > _but given the situation, How does that prevent the situation
         | from happening again_
         | 
         | You don't. Instead, you make sure your failover or DR setup is
         | regularly tested and works.
        
         | toast0 wrote:
         | Historically in the US, some portion of Bell installations were
         | designed to be resistant to attack. But it comes at large
         | expense for construction and maintenance. Underground
         | facilities also bring increased risk of flooding.
         | 
         | Competition and deregulation and lack of attacks leads towards
         | less robust installations to reduce costs. Geographically
         | redundant installations help as long as all installations
         | aren't targetted; and are valuable for operational concerns
         | other than just attacks.
        
           | userbinator wrote:
           | Cold War era definitely resulted in a lot of comms
           | infrastructure being hardened against attack.
        
           | riffic wrote:
           | not even a decade ago for an example:
           | 
           | https://en.wikipedia.org/wiki/2020_Nashville_bombing
        
         | SoftTalker wrote:
         | Data centers are usually built to withstand local natural risks
         | e.g. weather. All bets, SLAs, and insurance are usually off
         | when it comes to acts of war.
        
           | crote wrote:
           | There's also just an upper limit to the kind of risk you can
           | reasonably defend against.
           | 
           | An out-of-control wildfire levels the entire city? The Big
           | One hits the Bay Area? The entire city is flooded for a few
           | months because the levees break during a Cat5 hurricane?
           | Yeah, your DC will be completely ruined. And even if it
           | isn't, you're probably not getting any outside power,
           | generator fuel, or repair technicians for a while.
           | 
           | No matter how much money you pump into hardening your own
           | super-bunker DC, there will _always_ be disasters you aren 't
           | prepared for. At a certain point it just makes more financial
           | sense to abandon the idea of invulnerability and build a
           | redundant site a few states over. Accept that you _will_
           | occasionally lose one, and only protect against incidents
           | where mitigation is cheaper than occasionally rebuilding.
        
         | crote wrote:
         | > will there be a demand (all be small) for nuclear bunker
         | esque datacenters
         | 
         | Those already exist. See for example Bahnhof's "Pionen - White
         | Mountain" data center in Stockholm, or Cyberfort's "The Bunker"
         | a bit west of London.
        
       | bgentry wrote:
       | The important quote from the timeline:
       | 
       |  _Mar 01 9:41 AM PST
       | 
       | We want to provide some additional information on the power issue
       | in a single Availability Zone in the ME-CENTRAL-1 Region. At
       | around 4:30 AM PST, one of our Availability Zones (mec1-az2) was
       | impacted by objects that struck the data center, creating sparks
       | and fire. The fire department shut off power to the facility and
       | generators as they worked to put out the fire. We are still
       | awaiting permission to turn the power back on, and once we have,
       | we will ensure we restore power and connectivity safely. It will
       | take several hours to restore connectivity to the impacted AZ.
       | The other AZs in the region are functioning normally._
        
         | Imustaskforhelp wrote:
         | > we will ensure we restore power and connectivity safely
         | 
         | this would require human intervention and I am a bit worried
         | what if the strike can happen again and human lives might be
         | lost.
         | 
         | IIRC there have been cases in history where sometimes a same
         | location is targeted across multiple days. Obviously, AWS might
         | have local employees working in the region but would there be
         | an evaluation of this threat itself within the relevant team in
         | AWS. What if they try to bring the service back but then
         | missiles are struck again and what if human lives might be lost
         | on it. Let's just hope that it could be part of a evaluation as
         | well.
        
           | tokyobreakfast wrote:
           | > this would require human intervention
           | 
           | that's the difference between heroes and ordinary employees
           | who bitch about having to go into the office twice a month.
           | 
           | same as the stories you hear of guys taking snow-cats up a
           | mountain in a blizzard to restore phone circuits or radio
           | transmitters gone offline.
        
             | flymasterv wrote:
             | Man, don't be a "hero" trying to restore a lower ping to
             | someone trying to buy a kindle in Jeddah.
        
               | ok_dad wrote:
               | What about local hospitals which may have service from
               | that data center? There are heroes needed everywhere, all
               | the time.
        
               | stonogo wrote:
               | In that case, the hero was the person who avoided relying
               | on a single AZ when they deployed to cloud.
        
             | thatguy0900 wrote:
             | I'm sure bezos will be really happy someone is being a hero
             | for him in a war zone while he sails his newest yacht to
             | wherever the new version of the island is.
        
               | tokyobreakfast wrote:
               | on second thought there is a difference between restoring
               | critical infrastructure in times of crisis vs restoring
               | bot infrastructure for indian spamming operations. choose
               | wisely
        
           | pwarner wrote:
           | But I mean,are the employees safe at home? I guess if the
           | really targeted the data center then home is safer, but in
           | the fog of war maybe the data center wasn't the target?
        
         | jiggawatts wrote:
         | This reminds me of a visit to an Equinix data centre where the
         | sales person was droning on and on about how incredibly
         | reliable their power supplies were, how _uninterruptible_
         | everything was, etc, etc...
         | 
         | Essentially, he was trying to assure us that no-no-no, we don't
         | need multiple zones like the public clouds, they can instead
         | _guarantee_ 100% uninterrupted power under all circumstances.
         | 
         | A bit bored and annoyed, I pointed to the giant red button
         | conspicuously placed in the middle of a pillar and asked what
         | it is for.
         | 
         | "Oh, that's in case there's a fire!"
         | 
         | "What does it do?"
         | 
         | "It cuts... the power... uhh... for the safety of the fire
         | department."
         | 
         | "So... if there's a wisp of smoke in a corner somewhere, the
         | fireys turn up, the first thing they do is... cut the power?"
         | 
         | "... yes."
         | 
         | "Not 100% then, is it?"
        
           | vntok wrote:
           | Should have pushed it.
        
       ___________________________________________________________________
       (page generated 2026-03-01 23:00 UTC)