Yandex has finally given a number for what the October data-centre attacks permanently destroyed. In a statement on 10 November, the company said its engineers have restored access to 97.16% of customer data in Yandex Cloud. The remaining 2.84% cannot be recovered “due to critical damage to equipment and the absence of backups on the platform by customers”.
Reuters reports that three Yandex data centres were damaged in Ukrainian drone strikes in early October. Delovoy Peterburg names the sites as Sasovo (Ryazan region), Kaluga and Vladimir, with Sasovo and Vladimir fully stopped and several modules at Kaluga knocked out between 8 and 11 October.
The backup line is the real news
Yandex is careful to point out that customers who used Yandex Cloud’s own backup tools lost nothing: “all 100% preserved”. The losses fall on customers who didn’t. In response, Yandex Cloud has changed its default. Backup used to be something you switched on yourself. Now it is enabled by default on newly created virtual machines, on the same terms as before. You can still opt out, but Yandex calls it the “basic way to reduce risk”.
Yandex also says engineers are still spreading Cloud computing resources across “new independent locations in different regions of the Russian Federation”, and that the number of locations will keep growing. Delovoy Peterburg adds context: on 29 October Yandex said the B2B segment had been hit hardest, that customer data hadn’t been fully backed up, and that it had recovered 90% of lost data and capacity at that point. So the recovery has gone from 90% to 97.16% in under two weeks, and the remaining 2.84% has now been written off.
Why search people should care
Yandex Search tools sit on the same Cloud infrastructure. It’s the reason Search API has been running only at night, at 10–20% of its old capacity, since 1 November, and the reason rank-tracking vendors moved to live SERP parsing. It’s also the background to Webmaster’s slow, piecemeal restore and the indexing freeze that SEOs were still describing in yesterday’s forum recap.
The statement does not say anything specific about Search API capacity, Webmaster or crawling. It doesn’t say whether any of the 2.84% belonged to Yandex’s own services rather than external customers. Delovoy Peterburg says Yandex has restored “key functions of all services for users and business”, but the company has not put a date on full Search API daytime capacity. So don’t read “97% restored” as “Yandex is back to normal” for SEO tooling.
Our take
This is the first time Yandex has drawn a line under the data loss, and it confirms what agencies working in Russia suspected: anyone hosting crawlers, log stores or rank-tracking databases on Yandex Cloud without platform backups may have lost history for good. If that includes you or your vendors, find out now which date ranges are gone, before the year-end reports. Mark October 2026 as a break in the series in any long-run Yandex visibility data. And if you’re spinning up new infrastructure in Yandex Cloud, keep the new default backup switched on. It’s cheaper than explaining a gap to a client.
We’ll keep watching for the one update that would actually change things for SEOs: Search API back to daytime capacity, and Webmaster accepting new sites again.