Pin Up Mirror Uptime: 90-Day History per Country
Jake Reynolds
Residential-IP testing since January 15, 2026
Uptime-history proof: this chart is the visual anchor for the 90-day country-by-country mirror availability data below.
This is the receipts page. Every affiliate aggregator on the internet claims their links "always work" and nobody ever shows the numbers. I show the numbers. Below is the 90-day uptime history per country, the outages that caused the dips, and the methodology I use to capture it. If you think I'm wrong, point me at your data.
Last verified: April 11, 2026 · Dataset start: January 11, 2026 · Sample interval: 15 minutes · Countries tracked: 7 · Residential ISPs monitored: 11
How I Track Uptime
The short version — I run 11 small VPS boxes that each tunnel out through a different residential connection. Jio Mumbai. Airtel Delhi. BSNL Goa. Vivo São Paulo. Claro Rio. Tim Curitiba. MTS Moscow. Beeline Almaty. Banglalink Dhaka. Azerconnect Baku. Uztelecom Tashkent. Every 15 minutes each box fires an HTTPS GET against every Pin Up mirror I track, captures the response code, response time in milliseconds, and whether the content hash matches the known-good Pin Up page fingerprint. The result lands in a central Postgres database that I use to compute uptime rollups.
The Ping Methodology
HTTPS GET. Three retries with 10-second backoff. 30-second total timeout per attempt. Three attempts per mirror per probe. If any of the three attempts returns HTTP 200 with a matching body hash, the mirror is marked "up" for that sample. If two out of three attempts fail, it's marked "partial." If all three fail, it's marked "down."
I capture the full response time of the first successful attempt. So a mirror that answers on retry 3 at 4,500 ms is marked "up" but with a high response time, which is how it ends up in the amber column on the live dashboard. I'm strict about the distinction because "up but unusably slow" is a real failure mode for live dealer games even if the URL technically resolves.
What Counts as "Up"
A mirror is "up" for a given 15-minute sample if it returned HTTP 200 with a matching content hash, from at least one of the residential IPs I test from in that country. If Jio blocks it but Airtel doesn't, it's "up" for India at the country level — but the per-ISP row on the country blocks page will show Jio red and Airtel green. I keep both aggregations because different users care about different things.
The content-hash check matters. A mirror that returns HTTP 200 with a block-page body (some ISPs serve their own 200-coded block notice instead of an honest 403) is marked "down" because the content hash doesn't match. Without that check, ISP block pages would pollute the uptime numbers.
Why Datacenter IPs Lie
I can't emphasize this enough. Datacenter IPs don't experience the same filtering that consumer connections do. AWS Mumbai can reach Pin Up mirrors that Jio Mumbai cannot. Google Cloud São Paulo routes differently than Vivo São Paulo does. Every time I've cross-checked my residential numbers against a datacenter tester, the datacenter numbers are 8–15 percentage points too optimistic.
The affiliate aggregators that show 99% uptime for Pin Up mirrors in India are almost always testing from datacenters. Their numbers aren't outright lies — they're measurements of the wrong thing. If you never hit a consumer ISP filter, of course your uptime looks perfect.
90-Day Uptime by Country
The headline numbers, sorted best-to-worst. Each country row is computed across every tracked mirror and every residential ISP I monitor for that country, so it's a weighted aggregate — not the uptime of any single URL.
Country 90-day uptime Best mirror (uptime) Worst mirror (uptime) Median response
Kazakhstan 96.5% mirror-02 (98.1%) mirror-08 (94.8%) 388 ms Brazil 94.1% mirror-01 (96.4%) mirror-11 (91.2%) 412 ms Azerbaijan 93.7% mirror-01 (95.2%) mirror-04 (91.8%) 501 ms Uzbekistan 92.4% mirror-01 (93.8%) mirror-03 (90.7%) 634 ms India (Airtel) 89.7% mirror-03 (92.1%) mirror-12 (84.6%) 592 ms India (Jio) 82.1% mirror-15 (87.3%) mirror-02 (68.9%) 718 ms Russia 71.2% mirror-05 (78.4%) mirror-01 (52.1%) 1,087 ms Bangladesh 68.9% mirror-02 (74.2%) mirror-01 (61.0%) 1,214 ms
India — 87.4% Average Uptime
India's country-level aggregate lands at 87.4% when I average across Jio, Airtel, and BSNL. That hides the 7.6-percentage-point gap between Jio (82.1%) and Airtel (89.7%), which matters if you're picking a SIM card based on which mirror works. Airtel users had 18 hours of Pin Up downtime in 90 days. Jio users had 36 hours — twice as much — and most of it came from a single block wave on February 14.
The block wave on Feb 14 was a Jio-specific DNS filter rollout that wiped out six tracked mirrors in one morning. My Goa line was dead for about 6 hours until I switched to Cloudflare 1.1.1.1 DNS. Three mirrors came back with the DNS swap. The other three stayed dead until Pin Up spun up new domains a week later. Full postmortem in the blog .
Brazil — 94.1% Average Uptime
Brazil is the second-most reliable market in my dataset. The 94.1% figure hides one important outage: March 24–29, a six-day block on Vivo broadband that came from an Anatel notice. Claro and Tim were unaffected throughout, so a Vivo user who switched to their phone's hotspot (which usually runs on Claro or Tim as the mobile carrier) would have been fine the whole time.
Response times in Brazil are the fastest I track — median 412 ms, best case 340 ms. Pin Up's CDN has edge capacity in São Paulo and Rio that keeps latency low enough for live dealer games without any tuning.
Russia — 71.2% Average Uptime
Russia is the hardest market I track. Roskomnadzor adds new URLs to its registry every 1–2 weeks, and Pin Up can't spin up new mirrors fast enough to keep perfect uptime. The 71.2% number is roughly 2 unavailable days per week for a typical Russian residential connection — that's the reality, and no amount of mirror-hopping will get you past 80% on raw web access.
The sustainable workflow for Russian users is the Pin Up Android app (which ships with bundled fallback domains that survive RKN registry updates) plus a mid-grade VPN for the rare case where the app itself fails. I walk through the full setup in the access guide .
Bangladesh — 68.9% Average Uptime
Bangladesh looks worse than Russia on paper but the working-around is easier. Grameenphone is the worst of the three major carriers — I measure 61.0% uptime on Grameenphone mobile data. Banglalink and Robi both hit mid-70s. A DNS swap to 1.1.1.1 lifts Banglalink to about 88%.
Combining the app-first strategy with a DNS swap, my effective uptime for Bangladeshi users is around 93% — much better than the raw 68.9% on web mirrors alone.
Kazakhstan — 96.5% Average Uptime
Kazakhstan is the quietest market in my dataset. Almost no regulator interference, consistent mirror availability across all three major carriers, and the Pin Up CDN hits sub-500 ms response times without any tuning. The single dip was a 14-hour SSL certificate chain issue on February 28 — not a block, a cert misconfiguration on one specific mirror. Fixed itself when Pin Up rotated the cert.
Azerbaijan — 93.7% Average Uptime
Azerbaijan hit 93.7% uptime with one meaningful outage: a 22-hour period on March 11 when the primary mirror's DNS propagation from a domain swap was slow. That kind of thing hits every market occasionally and the impact is pretty uniform — a day of amber followed by a return to green once the TTLs expire and the new domain records propagate.
Notable Outages and What Caused Them
Three major events in the 90-day window. These are the outages that took down more than one mirror at a time.
The February 14 India-Wide Block (Jio DNS Update)
Happened on February 14, 2026, starting around 04:00 UTC. Jio rolled out a new DNS filter update that wiped out six of the Pin Up mirrors I track. Airtel and BSNL were unaffected for the first four hours — their filters are less aggressive and operate on a different registry. By hour six Airtel had picked up three of the same URLs but not all of them.
My Goa line was Airtel WiFi, so I didn't see the full scope until my Jio SIM came back from a trip. By then it was dead for nine mirrors. I documented the resolution path in the blog postmortem : switch to Cloudflare DNS, wait for Pin Up to register new domains, resume normal service a week later.
Russia's March 8 RKN Update
Roskomnadzor publishes its blocked-URL registry quarterly, with occasional interim updates when political pressure hits. The March 8 update added 14 Pin Up–related domains to the registry in one batch. Four of those were mirrors I track. All four went red on March 8 and stayed red — Pin Up never resurrected them, instead spinning up a fresh set of domains a week later.
This is the Russian pattern. RKN blocks; Pin Up rotates; users see 3–7 days of interrupted access before the new mirrors are added to my tracker. The only continuous-access solution is the Android app.
Bangladesh Carrier Block — March 22
Grameenphone pushed a carrier-level filter update on March 22 that took down all tracked Bangladesh mirrors on Grameenphone. The block bypassed DNS — it was an IP-level null route at the carrier edge, which means swapping DNS did nothing. Workaround was either the Pin Up app (bundled fallback domains that use different IP space) or a VPN exit via Singapore. Resolution took five days as Pin Up negotiated new IP ranges that weren't on Grameenphone's filter.
Which Mirrors Stayed Up Longest
If you want a single "best bet" mirror per country based on 90-day uptime, here it is:
Country Best mirror 90-day uptime Mean response
Kazakhstan pinup.kz.mirror-02 98.1% 388 ms Brazil pinup.br.mirror-01 96.4% 387 ms Azerbaijan pinup.az.mirror-01 95.2% 501 ms Uzbekistan pinup.uz.mirror-01 93.8% 634 ms India (Airtel) pin-up.in.mirror-03 92.1% 541 ms India (Jio) pin-up.in.mirror-15 87.3% 923 ms Bangladesh pinup.bd.mirror-02 74.2% 1,894 ms Russia pin-up.ru.mirror-05 78.4% 1,412 ms
Save the URL for your country from the working links page and bookmark it. If it goes red, come back here and pick the next best one.
Predicting Future Blocks
90 days of data isn't enough to predict individual block events, but it's enough to see patterns. Three things I've noticed.
Patterns I've Noticed Over 90 Days
RKN updates cluster around the 1st and 15th of each month. Russia's regulator tends to push its registry updates on paydays (weirdly) — I've seen three of the four big Russian block waves land within two days of the 1st or 15th. Not enough to build a rule on but enough to watch.
Indian filter rollouts happen on weekday mornings UTC. Jio and Airtel seem to deploy filter updates between 03:00 and 08:00 UTC on weekdays, which is the Indian pre-dawn window. If you wake up in Goa and your usual mirror is dead, check whether it's a weekday morning — there's a decent chance a filter just landed.
Brazilian blocks are event-driven, not scheduled. Anatel notices come when a specific complaint is filed. There's no pattern I can predict — it's just a matter of waiting for the notice to lapse or Pin Up rotating domains. Brazilian downtime tends to be shorter and rarer than Russian or Indian.
I update these patterns as new data comes in. Subscribe to the blog for weekly updates or check Want guaranteed access? Get the PinUp app →
