The Squid’s InkSquidGrid’s weekly read of the internet — fresh ink every Monday

Week 42, 2026 · Oct 5 – Oct 11

Covers Monday Oct 5 00:00 to Monday Oct 12 00:00 UTC.

How to read this: we measure millions of internet routes around the world. Every speed number below is how many milliseconds (thousandths of a second) slower a route ran than its own normal speed, so higher is worse. Each section says whether things got better or worse than last week.

2,495,662internet routes we measure
626,879,614measurements taken this week down 9% from last week
1,495outages and problems companies reported down 1% from last week
530of them rated major or critical

SG.GS (Singapore network operator)’s rough weekthe week’s story

About the same as last week. SG.GS (Singapore network operator) is one of the large networks that carry internet traffic between countries. Routes through it ran 46.1 milliseconds slower than normal on average this week (last week: 46.6). 855 routes through it were badly slowed (more than 25 milliseconds over normal) at some point. The chart shows how many, hour by hour:

Signed-in customers see the affected routes, the carrier on each side of every changed handoff, and the regional split on the SG.GS (Singapore network operator) assessment page.

The networks that slowed down the most

Large carrier networks, ranked by how much slower than normal their routes ran this week. Big networks get no allowance for their size.

Networkthis week, hour by hourslower than normalroutes badly slowedlast weekcompared with last week
SG.GS (Singapore network operator)46.1 ms85546.6 msAbout the same as last week
FLAG Telecom36.4 ms26337.4 msBetter than last week
Bharti Airtel Limited30.3 ms1,70730.6 msAbout the same as last week
TransTelecom26.3 ms53626.2 msAbout the same as last week
Vocus Communications26.3 ms411—New to this list this week
SEACOM26.0 ms41226.8 msAbout the same as last week

“Routes badly slowed” counts routes that ran more than 25 milliseconds over normal at some point this week.

1,495 outages and problems reported by companies

Fewer than last week (down 1%). These are problems online services posted on their own status pages. Last week: 1,513.

SaaS895
Payments & fintech235
AI113
Crypto45
Cloud & hosting44
Consumer33
Education & research29
Gaming26

Online business software (SaaS) had the most problems again, as it usually does.

What tended to break, in providers’ own words: connectivity ×146 · slowness ×139 · elevated errors ×97 · auth & login ×68

Longest problems to fix: Spendesk — Amazon Payments – Card Tokenisation Failure (498 hours, about 21 days); Sprinklr — Planned Infrastructure Enhancement Activity for Prod15 (490 hours, about 20 days); OSG Consortium — Nebraska Data Center Move (484 hours, about 20 days).

Ground eventsthe physical world, and whether the internet flinched

2026-10-08 Russia: drone strike knocks a Yandex data center offline
Overnight on October 8, Ukrainian drones hit a data center in Sasovo, in Russia's Ryazan region. The data center belongs to Yandex, one of Russia's biggest tech companies. The strike started a fire. No one was hurt, but the site shut down completely for a time, Meduza reported. Some Yandex services went down, and so did services of other companies that ran on the same site. The monitoring group NetBlocks confirmed a major drop in connectivity on Yandex's network. A later report citing Cloudflare data said more Yandex sites were hit through October 11, and that Yandex warned customers it could not say when all services would come back.
Meduza: Ukrainian drones struck Yandex's Sasovo data center on Oct 8, causing a fire and a temporary full shutdown · NetBlocks (Mastodon): metrics confirm a significant disruption to Yandex's network after the Sasovo fire · NetBlocks (X): same confirmation of the Yandex network disruption, Oct 8 · OffSeq Radar: report citing Cloudflare data says three Yandex data centers were hit in four days; Yandex warned of outages
2026-10-09 India: mobile internet cut in central New Delhi for 24 hours
India's Home Ministry told mobile phone companies to switch off mobile internet within 4 kilometres of the Janpath and Kartavya Path crossing in New Delhi. The cut ran from 10 pm on October 9 to 10 pm on October 10. It came ahead of a planned protest at Jantar Mantar by the Cockroach Janta Party and others. Police sources said internet in the rest of Delhi was not affected. It was the second cut in the area in eight days. On October 2, mobile internet was off for 12 hours in a smaller area around Jantar Mantar. MediaNama reported that the order was marked "top secret" and was dated the same day the cut began, which left people very little warning.
Business Standard (PTI): mobile internet suspended within 4 km of Jantar Mantar, 10 pm Oct 9 to 10 pm Oct 10; rest of Delhi unaffected · MediaNama: 'top secret' Home Ministry order; earlier 12-hour, 1.5 km shutdown on Oct 2 · Business Today: 24-hour mobile internet block ahead of the CJP rally at Jantar Mantar on Oct 10

Ongoing, week 17 of tracking: Iran, slightly worse than last week. The months-long shutdown that began in January ended on May 26, but heavy limits on internet use remain. In September, lawmakers drafted a new bill that critics say would tighten online censorship even more, RFE/RL reported. Routes into Iran (57 we measure) ran 1.0 milliseconds slower than normal, up from 0.7 last week.

Ongoing, week 2 of tracking: Philippines cable cut, better than last week, according to the phone and internet companies. PLDT and Globe said services were back to normal by October 4 after they moved traffic onto other links. The broken cable itself is not fixed yet. Globe said that repair will take several weeks. We still have no measurements for the Philippines, so we cannot compare with our own numbers.

Ongoing, week 1 of tracking: Russia data center strikes, new this week. Ukrainian drones hit a Yandex data center in Sasovo on October 8, and one report said more Yandex sites were hit through October 11. Routes into Russia (9,099 we measure) ran 11.5 milliseconds slower than normal. That is worse than last week's 10.8. Our numbers alone cannot tell us whether the strikes caused the change.

We cover earthquakes of magnitude 7.0 or more and storms at the top two disaster alert levels. “Our fleet’s view” is what our own measurements saw, not a damage report.

Subsea watchwhere the internet is wet

2026-10-04 Philippines–Singapore cable: service back, repair still to come
An update on last week's cable cut. PLDT said its internet service was back to normal across the country by the afternoon of Sunday, October 4. Globe also said it had restored service by moving traffic onto other international links. Globe added that fixing the damaged cable itself will take several weeks, and that it is still working with partners to confirm what caused the damage. On October 6, Converge said its newer undersea cables to Singapore kept its customers online through the disruption.
Inquirer: PLDT services fully restored nationwide, 4:30 pm Sunday · BusinessWorld: Globe reroutes traffic; cable repair to take several weeks; cause still being checked · Philstar: Converge says its newer subsea cables spared its customers from the outage

Also on the wire this week: SUBCO Weighs Australian Cable Ship and Survey USV (2026-10-07); Philippines-Singapore C2C Link Restored (2026-10-05); Philippines Cable Disruption Slows Internet (2026-10-05).

China’s internet gateways

China’s three big carriers control most traffic in and out of the country. We check how much of each network we can reach from outside China. The percentage is how much we reached; “normal” is its usual level. Lower means more is blocked or unreachable.

China Telecom 54.3% reachable (normal 52.4%)
About normal
China Mobile 79.1% reachable (normal 82.5%)
Worse than normal
China Unicom 82.5% reachable (normal 83.6%)
About normal

“About normal” means within 3 percentage points of the usual level.

For the feedpaste-ready

The Squid's Ink -- Week 42 This week, war reached the server room. Overnight on October 8, Ukrainian drones hit a Yandex data center in Sasovo, Russia. The strike started a fire and shut the site down for a time. Some Yandex services went offline, and so did services of other companies that relied on that site. NetBlocks confirmed a major drop on Yandex's network. One later report said more Yandex sites were hit through October 11. Our own numbers show routes into Russia ran 11.5 ms slower than normal, up from 10.8 ms last week. We can't say from our data alone whether the strikes caused that. Also this week: a 24-hour mobile internet cut in central New Delhi, and the Philippines is back online while it waits weeks for a cable repair. #InternetHealth #TheSquidsInk

Every figure derives from SquidGrid’s own measurement fleet and its archive of provider-published incidents; provider self-reports are labeled as such. Provider Assessment Reports carry the per-provider detail. The internet was mostly fine this week. We measured anyway.