Internet Resilience in Q3 2024: A Quarter of Multi-Faceted Outages
Cloudflare’s network, which interconnects with over 13,000 providers across more than 330 cities in 120+ countries, offers a unique vantage point for observing Internet resilience. Using the routing and traffic analysis features released on Cloudflare Radar earlier this year, we can now assess the impact of disruptions at both network and location levels. This summary highlights significant, confirmed outages; it is not exhaustive, and a more complete list of anomalous traffic events is available in the Cloudflare Radar Outage Center.
The third quarter of 2024 was exceptionally disruptive. Government-directed shutdowns during exam periods remained a recurring theme. Physical infrastructure damage—both terrestrial and submarine cable cuts—severely hampered connectivity in Africa and elsewhere. An unusually active hurricane season produced widespread outages across the Caribbean and parts of the United States. In South America and Africa, electricity supply issues, both scheduled and unscheduled, led to multi-hour connectivity losses. Military operations, infrastructure maintenance, a fire, and a suspected cyberattack also contributed to the quarter’s instability, alongside several notable incidents with no confirmed root cause.
Government-Directed Shutdowns During Exams
Following a pattern established throughout 2024, several nations imposed Internet shutdowns to curb cheating during national examinations. These actions caused sharp and deliberate traffic drops across multiple regions.
- Mauritania: A country-wide shutdown occurred for roughly three hours on October 2, from 08:00 to 11:00 UTC, coinciding with the baccalauréat exam period.
- Liberia: The government ordered a nationwide shutdown from 09:00 to 12:00 UTC on October 3 during the West African Senior School Certificate Examination (WASSCE) roll-out.
- Iraq: A nationwide shutdown was implemented on October 12 from 16:00 to 19:00 UTC, and a follow-up regional action occurred on October 15 in the Kurdistan Region from 16:00 to 19:00 UTC.
The traffic impacts were stark. In Mauritania, traffic fell to zero during the window before recovering. Similar zero-traffic plateaus were observed in the affected regions of Iraq during the mandated hours.
Major Cable Cuts
Damage to both submarine and terrestrial cables caused prolonged connectivity loss. The incidents underscored the fragility of specific cable routes and the lack of redundant paths in some regions.
On September 22, Azerbaijan faced a major disruption beginning at approximately 15:30 UTC. The outage originated in the western direction due to a cut on the Caucasus Online cable system between Georgia and Azerbaijan. Traffic from Azerbaijani networks, including AS29124 (Delta Telecom) and AS28797 (Azerconnect), was observed being locally rerouted to Russian providers. International connectivity remained reduced for more than six hours.
In Africa, suspected sabotage in the Red Sea and a cut in Morocco affected WIO, SEACOM, and other cable systems on July 14, impacting carriers like AS24863 (WIOCC) and AS37100 (SEACOM). Several submarine cables also broke in Côte d’Ivoire over the quarter, leading to a series of disruptions.
A less common terrestrial issue arose in Vietnam on September 22, where a pair of cuts on the Asia Africa Europe 1 (AAE-1) cable system and the APG submarine cable caused significant packet loss and latency increases for carriers like AS18403 (FPT Telecom) and AS7552 (Viettel). In China, multiple failures on the SEA-ME-WE 4 cable during September resulted in periods of degraded connectivity for networks such as AS9808 (China Mobile).
Hurricane and Weather Impact
The 2024 Atlantic hurricane season was remarkably active, with several systems directly hitting land and causing significant Internet infrastructure damage. The most substantial impacts began in late September with Hurricane Helene.
- Hurricane Helene (landfall September 26): Made landfall in Florida’s Big Bend region. While Florida sustained initial impacts, catastrophic flooding in western North Carolina caused severe, prolonged outages across multiple counties. Buncombe County, home to Asheville, saw traffic drop to near zero for several days. Other affected areas included Tennessee, Georgia, and Virginia, with emergency response networks also reported as degraded.
- Hurricane John (landfall September 24): The storm significantly impacted Acapulco, Mexico, causing a traffic drop to near-zero levels for several hours on September 25-26.
- Hurricane Francine (landfall September 11): In Louisiana, the hurricane and resultant flooding caused a notable but recoverable traffic decline in the New Orleans area.
- Late September – October: Hurricane Isaac brought disruptions to Martinique and Guadeloupe, while Hurricane Kirk caused minor issues in the Azores.
The storms in the Caribbean, particularly Beryl, Ernesto, and others, caused periodic but short-lived disruptions across islands like Jamaica, Grenada, and the Bahamas. While traffic often rebounded, intermittent connectivity issues continued as a result of power fluctuations and damaged backhaul links.
Hurricane Helene and the subsequent flooding in North Carolina proved to be one of the most catastrophic Internet resilience events of the year, with recovery taking weeks for many isolated communities. Furthermore, on October 1, a cellular network failure in parts of Mexico, which a major operator initially attributed to cable cuts, was later reported by the local regulator to have been caused by a software bug. This highlights the complexity of diagnosing post-event network failures.
Power Outages and Other Operational Failures
In South America, two of the largest economies faced widespread connectivity losses due to electricity issues. In Argentina, a high-voltage transmission line failure near Buenos Aires on October 6 led to a nationwide blackout starting around 15:45 UTC. Traffic from major national networks dropped by 40-60% immediately. Power was restored in waves over the following hours. In Brazil on October 15, a failure at the Itaipu hydroelectric plant caused a massive power outage affecting 26 states. Major backbone providers AS28573 (Claro) and AS18881 (Vivo) saw traffic drop by nearly 100% in the immediate aftermath, with many national networks taking hours to recover.
African nations experienced similar issues. A fire at Eskom’s Medupi power station on October 2 in South Africa forced load-shedding, degrading connectivity. In Nigeria, multiple power grid collapses on September 15 and October 7 led to reduced traffic, particularly affecting the northern states.
Across Europe, government-mandated IPv6 transition tests in Switzerland (September 25) and Germany (October 2) caused minor, localized disruption as carriers participated in the tests. Meanwhile, non-weather-related fibre cuts in Spain, Denmark, Ireland, and Portugal occasionally occurred during maintenance windows, but these were largely localized. Maintenance actions affecting AS7029 (Windstream) in the US and AS1239 (Lumen) were observed but followed expected patterns.
Incidents Without Complete Information
Several major incidents during the quarter lacked verified or singular root causes. On July 20, a nationwide disruption in Tunisia caused a significant drop in traffic, though the cause remains unclear despite the government conducting an investigation. In Ukraine, a nationwide decline in Internet traffic was observed on September 25-26, reported as a result of a suspected cyberattack on a major telecom infrastructure provider, although specific details were never fully confirmed to align with all anomalies. These events highlight the ongoing challenge of identifying the precise reasons behind Internet outages in complex operational environments.
Bangladesh’s escalating crackdowns
Violent student protests against government job quotas and unemployment prompted Bangladesh’s government to order a nationwide mobile Internet shutdown on July 18, reportedly to “ensure the security of citizens.” The disruption escalated into a near-complete national outage as broadband networks were also taken offline. According to Cloudflare Radar data, country-level Internet traffic and announced IP address space dropped to near zero just before 21:00 local time (15:00 UTC).
Individual providers went down in stages that same day. Complete outages, marked by simultaneous losses of traffic and announced IP address space, hit AS24389 (Grameenphone) at 01:30 local time and AS25245 (Banglalink) at 02:15 local time. AS24432 (Robi Axiata) followed at around 06:30 local time. AS58715 (Earth Telecommunication) saw traffic begin falling at 18:00 local time and reach zero by 22:00, with announced address space fully withdrawn by 21:25. AS63526 (Carnival Internet) lost traffic at 20:45 local time and had its address space completely withdrawn within the following hour.
Fixed broadband connectivity began recovering around 18:00 local time on July 23, with mobile networks following on July 28. According to the state minister for post, telecommunication and information technology, the initial restoration was a “trial run” prioritizing banking, commercial sectors, technology firms, exporters, outsourcing providers, and media outlets.
Renewed clashes between police and protestors on August 4 led the government to again order mobile Internet shutoffs, with traffic on the mobile providers dropping between 13:30 and 14:15 local time (07:30 to 08:15 UTC). By August 5, the government escalated to a full national Internet shutdown, with traffic and announced address space falling sharply around 10:30 local time (04:30 UTC). This one was short-lived: broadband began recovering around 13:20 local time (07:20 UTC), and mobile connectivity was restored around 14:00 local time (08:00 UTC).
Exam-season shutdowns in Iraq and Iraqi Kurdistan
Iraq and the autonomous Kurdistan region in its north regularly impose government-directed Internet shutdowns during secondary and baccalaureate exams. In Q3, Iraqi Kurdistan saw two rounds of such disruptions.
First, in July, AS59625 (KorekTel), AS21277 (Newroz Telecom), AS48492 (IQ Online), and AS206206 (KNET) were impacted between 06:00–08:00 local time (03:00–05:00 UTC) on July 3, 7, 10, and 14. These matched the pattern seen in the second quarter and in June 2023. None of the affected networks saw a drop in announced IP address space during these shutdowns, indicating the providers were offline rather than disconnected from the Internet.
A second set occurred across the back half of August — on the 17, 19, 21, 24, 26, 28, and 31 — with the same four providers affected during the same morning window, as traffic dropped between 06:00 and 08:00 local time.
In Iraq proper, a second round of exams for 12th graders led to more than two weeks of regular shutdowns. Between August 29 and September 16, disruptions occurred between 06:00–08:00 local time (03:00–05:00 UTC) on multiple days, with both HTTP traffic and announced IP address space dropping at the country level. (A red annotation on September 11–12 in the corresponding graphs reflects an internal data pipeline issue, not an outage.)
The affected providers included AS58322 (Halasat), AS51684 (AsiaCell), AS203214 (HulumTele), AS199739 (Earthlink), and AS59588 (ZAINAS). Distribution of mobile and desktop traffic showed gaps on AS58322 and AS199739 — and to a lesser extent AS203214 — during shutdown periods, suggesting these networks went fully offline. Meanwhile, AS51684 and AS59588 remained at least partially online, corroborated by their only partial loss of announced IP address space.
Syria and Mauritania follow the exam pattern
A second round of exam-related shutdowns in Syria, requested by the Ministry of Education, ran from July 25 through August 8 per a schedule published by Syrian Telecom on Facebook. Unlike the earlier round in May–June (covered in a separate post), these disruptions all began at 07:00 local time (04:00 UTC), with end times varying between 09:45 and 10:30 local time. The impact was visible at the country level and on AS29256 (Syrian Telecom), the country’s primary telecommunications provider.
In Mauritania, a round of baccalaureate exams that began August 12 triggered a series of shutdowns aimed at preventing cheating. Two occurred on August 12, from 08:00–12:00 and 15:00–19:00 local time (UTC), with a third the following day from 08:00–12:30 local time. The affected networks were AS37508 (Mattel), AS37541 (Chinguitel), and AS29544 (Mauritel). Announced IP address space for these networks stayed unchanged during the shutdowns, indicating that mobile subscriber connectivity was disabled rather than the networks being cut off from the Internet. Mauritania has a history of similar exam-related disruptions, having implemented them between 2017 and 2020.
When the fiber goes down
Fire, sanitation, and backhoe incidents in Africa
A series of fiber cuts across Africa in Q3 produced disruptions ranging from a few hours to several days. In Eswatini, MTN Eswatini (AS327765) reported intermittent connectivity on July 14 caused by fiber breaks from wildfires. Cloudflare Radar shows the effect between 19:30 and 20:15 local time (17:30–18:15 UTC).
In Cameroon, sanitation work in Yaoundé cut a fiber on August 4, taking down voice and data services for Cameroon Telecommunications (AS15964, Camtel) customers. The disruption ran from 03:00 to 16:30 local time (02:00–15:30 UTC). Although it began during a low-traffic overnight period, both request and bytes traffic stayed below the prior week's levels for the full duration.
Liberia's ACE submarine cable, which links West African countries to Portugal and Europe, suffered an interruption first detected at 01:00 local time on August 21, according to the Liberia Telecommunications Authority. Lonestar Cell MTN (AS37410) was among the providers "gravely affected." The network saw a complete outage lasting more than a day and a half, with traffic recovering at 14:00 local time on August 22. The operator later confirmed restoration and said customer accounts would receive 500 MB of free data.
Airtel Niger warned customers on September 7 of outages from international fiber cuts. As a landlocked country, Niger depends on terrestrial links to neighboring networks. Two disruptions were observed around that time: one from 18:00–21:00 local time (17:00–20:00 UTC) on September 6 and another from 10:45–12:00 local time (09:45–11:00 UTC) on September 7. Both were visible at country level and on AS37531 (Airtel Niger) and AS37233 (Orange Niger / Zamani Telecom).
Denied access extends a Haitian outage
Submarine cable failures usually take time to fix because the damage is hard to reach. A September 14 failure affecting Digicel Haiti (AS27653) dragged on for over a week for a different reason: technicians were refused access to the damage site.
The traffic loss began at 08:00 local time (12:00 UTC) on September 14. Two days later, Digicel confirmed a failure on an international submarine cable belonging to Cable and Wireless, with damage at Kaliko Beach Club, reportedly used as a cable entry point. Technicians were denied access to the scene due to a business dispute dating to 2021. Temporary measures kept essential services running, preventing a total blackout. A September 22 press release announced restored Internet service as of 02:00 local time (06:00 UTC) and cited vandalism as the cause.
A short, sharp outage in Kyrgyzstan
Damage to the backbone wire of an upstream provider left Kyrgyzstan's Megacom (AS50223) offline for just one hour on September 25, between 15:45 and 16:45 local time (09:45–10:45 UTC). AS12389 (Rostelecom) is listed as Megacom's only upstream provider. Both traffic and announced IP address space dropped to zero, and country-level traffic fell as much as 72% compared with the prior week. The complete loss suggests the damage occurred on the link between Megacom and Rostelecom.
Hurricanes and floods hammer the Caribbean and US Southeast
The third quarter brought an unusually active hurricane season that left a visible mark on Internet connectivity across the Caribbean and the southeastern United States. Beyond the storms, severe flooding in Nepal also knocked out service for several major providers.
Caribbean islands take the first hit
Hurricane Beryl tore through the eastern Caribbean at the start of July, causing widespread destruction in Grenada and Saint Vincent and the Grenadines. In Grenada, traffic began falling sharply at 10:00 local time (14:00 UTC) on July 1, just before landfall on Carriacou Island. The worst of the outage stretched over roughly the first 24 hours, with traffic not recovering to expected pre-storm levels until 10:00 local time (14:00 UTC) on July 5.
Saint Vincent and the Grenadines saw a similar pattern, with traffic dropping at the same time on the same day. The most significant loss occurred in the first 24 hours followed by a gradual recovery, but service did not normalize until July 11.
Jamaica felt Beryl's effects on July 3, when the storm passed the island. Traffic fell sharply around 14:00 local time (19:00 UTC), after a period of higher than normal volumes that likely reflected residents seeking storm information. The disruption lasted nearly a week, with traffic returning to expected levels on July 10.
In August, Tropical Storm Ernesto triggered power outages in the U.S. Virgin Islands that degraded Internet connectivity. Traffic dropped steeply at 22:00 local time on August 13 (02:00 UTC on August 14) and stayed low for over two days, recovering around 11:00 local time (15:00 UTC) on August 16.
Ernesto later strengthened into a hurricane but had weakened by the time it reached Bermuda on August 16/17. Damage was largely confined to power outages, downed trees, and flooding, yet even that was enough to disrupt connectivity. Traffic fell over 80% at 22:00 local time on August 16 (01:00 UTC on August 17) and remained depressed for roughly two and a half days, returning to expected levels around 09:00 local time (12:00 UTC) on August 19.
Helene's path through the Southeast
Hurricane Helene made landfall in northern Florida as a Category 4 storm late on September 26 and pushed north through Georgia, South Carolina, North Carolina, and into Tennessee. The storm caused historic flooding and severe damage to roads, homes, power lines, and telecommunications gear. State-level traffic data shows the scale of the impact.
In Georgia, Helene arrived early on the morning of September 27, and by midday peak traffic was roughly 20% lower than in the days before the storm. Peak traffic stayed depressed through the following week, with recovery becoming more apparent starting the week of October 6.
The most heavily impacted network in Georgia was AS11240 (ATC Broadband). Its traffic began dropping around 22:00 local time on September 26 (02:00 UTC on September 27) and customers faced a near complete outage until around 08:00 local time on September 30 (12:00 UTC), when volumes slowly recovered. Other Georgia providers hit hard included AS400511 (Clearwave Fiber), AS394473 (Brantley Telephone Company), AS40285 (Northland Cable Television), AS15313 (Pembroke Telephone Company), and AS397118 (Glenwood Telephone Company).
South Carolina saw its midday traffic peak on September 27 reach just 65% of the preceding days. Peaks remained lower over the weekend and the following week, with improvement visible the week of October 6.
At AS19212 (Piedmont Rural Telephone), traffic fell rapidly around midnight local time on September 27 (04:00 UTC) and reached near complete outage over the next eight hours. Recovery was gradual over several days, with a regular pattern emerging on October 1 and rapid growth through the following week. Other affected South Carolina networks include AS397068 (Carolina Connect), AS10279 (West Carolina Communications), AS20222 and AS21898 (TruVista), and AS14615 (Rock Hill Telephone).
North Carolina showed a different pattern. The traffic drop on September 27 occurred after a midday peak in line with previous days, and the magnitude was smaller than in South Carolina or Georgia. Traffic peaks over the following week stayed in line with the week before Helene, with higher peaks starting the week of October 6.
North Carolina providers AS53488 (Morris Broadband) and AS53274 (Skyrunner) both suffered multi-day disruptions likely tied to storm damage. Morris Broadband went completely offline several times over the course of a week — announced IP address space dropped to zero on three separate occasions, aligning with outages visible in traffic graphs. Skyrunner's outages were less severe but still significant: it lost 75-80% of its announced IP address space for a two-day stretch covering September 27-29. AS22191 (Wilkes Communications) and AS23118 (Skyline Telephone) also saw major impacts.
Texas outage and Nepal flooding
Hurricane Beryl also caused indirect disruption in Texas. AS11427 (Charter Communications/Spectrum) saw traffic drop between 12:30 and 19:30 local time on July 9 (17:30 to 00:30 UTC). The provider attributed the issue to a "third-party infrastructure issue caused by the impact of Hurricane Beryl" in a post on X, acknowledging the problem shortly after it began and confirming restoration after service returned.
Heavy rains in late September triggered flooding and landslides across much of Nepal, leading to power outages and Internet disruptions. On September 28, four major providers saw traffic fall 50-70% between 14:15 and 16:00 local time (08:30 to 10:15 UTC): AS23752 (Nepal Telecom), AS45650 (Vianet), AS139922 (Dishhome), and AS17501 (Worldlink).
Extended outages: power, maintenance, and conflict
Power grid failures
Several countries experienced Internet disruptions in Q3 2024 caused by failures in their national power grids. The causes ranged from infrastructure faults to planned maintenance, but each resulted in measurable dips in country-level traffic.
Venezuela: A nationwide power outage on August 30, which President Nicolás Maduro attributed to an attack on the Guri Reservoir hydroelectric project, caused country-level traffic to drop 82%. All 24 states reported total or partial electricity loss. Traffic began recovering as power was restored throughout the day, returning to expected levels just after midnight local time on August 31.
Kenya: Kenya suffered two major power outages in as many weeks. The first, on August 30, began roughly half an hour before Kenya Power Care issued a customer alert at 21:57 local time. Traffic fell as much as 61% and did not recover to expected levels until 06:00 local time the following morning, several hours after the utility reported power had been restored to affected regions.
A second outage struck on September 6. Traffic began dropping around 09:00 local time, falling approximately 45%. Energy and Petroleum Cabinet Secretary Opiyo Wandayi explained that the 220kV Loiyangalani transmission line tripped at Suswa substation while evacuating 288MW from Lake Turkana Wind Power, followed by a trip on the Ethiopia–Kenya 500kV DC interconnector carrying 200MW — a total loss of 488MW. Traffic returned to expected levels around 16:00 local time, aligning with the utility's announcement that normal supply had been restored at 3:49pm.
Ecuador: A fault in the Molino substation bar caused a power outage on September 7 in several provinces. Ecuador's national electricity operator CENACE predicted restoration within 3 to 4 hours; traffic returned to expected levels just under four hours after the drop began, matching the forecast.
A series of planned nightly maintenance outages followed on September 18/19, causing traffic to drop over 60% as compared to the prior week. The cuts were reportedly scheduled for 22:00–06:00 local time, and traffic recovered around 06:00. Further outages planned for September 23–27 did not appear to impact traffic levels relative to the previous week.
Senegal: An incident at the Hann substation on September 12 caused the loss of the OMVS interconnected network, according to power company Senelec. Internet traffic dropped sharply at 13:00 local time, falling as much as 80%. Traffic recovered to expected levels by 20:00, around the same time Senelec announced full restoration of electricity supply.
Maintenance incident: Syria
Syria had already seen an exam-related shutdown on July 30 from 07:00–10:15 local time. Just an hour after connectivity was restored, another disruption began. Syrian Telecom attributed it to an explosion during periodic maintenance of an air conditioner in a technical hall, which temporarily knocked Internet circuits out of service. Traffic remained depressed for approximately eight hours, recovering around 19:00 local time.
Cyberattack: Russia
Roskomnadzor, Russia's Internet regulator, blamed a brief disruption on August 21 on a distributed denial-of-service (DDoS) attack targeting Russian telecommunications operators. The disruption on AS12389 (Rostelecom) lasted from roughly 13:45 to 14:30 Moscow time. Roskomnadzor stated that the attack had been repelled by 3 PM. The outage reportedly impacted Telegram, WhatsApp, Wikipedia, Yandex, VKontakte, telecom support services, and mobile banking apps. Some experts questioned the official explanation, suggesting instead that centralized interference by Roskomnadzor was responsible.
Conflict-related disruptions
Palestine: On September 8, Paltel announced the suspension of home Internet services in central and southern areas of the Gaza Strip due to ongoing aggression. Traffic in the Gaza, Rafah, and Deir al-Balah Governorates dropped sharply at 18:00 local time, with the most significant impact in Rafah and Deir al-Balah. Services were restored around 23:00 local time, and Paltel confirmed the return of home Internet services in a subsequent post.
Lebanon: Israeli airstrikes targeting Beirut on September 28 likely knocked local network provider Solidere (AS42852) offline. Traffic loss began around 12:15 local time, coinciding with a complete loss of announced IP address space. Most address space was re-announced at 14:45, with a slight traffic increase at that time. Traffic fully recovered just after 18:00 local time.
Fire: Algeria
A fire near a data center in Blida Province disrupted connectivity on AS327931 (Djezzy) on July 24. The blaze broke out in a warehouse of a company located near Djezzy's technical center. The operator predicted that 97% of sites would be restored by around 3 am the following day, but traffic did not return to expected levels until the end of the day on July 25.
Nationwide mobile failures dominate the quarter's final day
United States
On September 30, Verizon mobile customers across multiple US cities reported losing connectivity. Phones displayed "SOS" instead of signal bars, and users could not make or receive calls. Reports began around 09:00 ET (13:00 UTC), though Cloudflare Radar only saw a noticeable change in request volume at the ASN level about two hours later. Verizon's mobile network operates under AS6167 (CELLCO).
Verizon acknowledged the issue shortly before 12:00 ET (16:00 UTC), stating that engineers were engaged. HTTP traffic dipped approximately 5% compared to the prior week around 11:00 ET (15:00 UTC), and fell as much as 9% below expected levels at 13:45 ET (17:45 UTC). Affected cities included Chicago, Indianapolis, New York City, Atlanta, Cincinnati, Omaha, Phoenix, Denver, Minneapolis, Seattle, Los Angeles, and Las Vegas.
Traffic normalized around 17:15 ET (21:15 UTC). At 19:18 ET (23:18 UTC), Verizon reported that service had fully returned to normal levels.
Pakistan
A roughly two-hour disruption hit Pakistan on July 31, between 13:30 and 15:30 local time (08:30–10:30 UTC). Country-level traffic dipped about 45%, but AS17557 (PTCL) saw a near-total loss of traffic, and AS24499 (Telenor Pakistan) dropped nearly 90%. Together, these two providers serve an estimated nine million users and rank among the country's top five ISPs.
The cause remains disputed. The Pakistan Telecommunication Authority reportedly blamed a technical glitch in an international submarine cable affecting PTCL's network, while other sources claimed a misconfigured government firewall intended for content blocking caused the disruption.
United Kingdom
On August 14, UK subscribers of Vodafone (AS25135) reported problems with mobile and landline Internet access. Traffic began declining around 11:00 local time (10:00 UTC), ultimately falling 43% below the prior week's levels. Service recovered by 13:30 local time (12:30 UTC). Vodafone did not publicly acknowledge the issue or provide an explanation.
Power infrastructure remains a weak link
The third quarter's disruptions stemmed from a variety of causes, but outages tied to aging or poorly maintained electrical infrastructure stand out. Losing power increasingly means losing Internet connectivity, as communication, entertainment, and commerce have become heavily reliant on the network. Mobile connectivity is not a complete substitute, particularly since devices eventually need recharging. Addressing the underlying infrastructure will require significant time, resources, and investment, though governments appear to be taking steps in that direction.
