We are troubleshooting an issue with the 2022Q2 realm that causes sites to intermittently become unavailable. If you are affected, you may wish to upgrade your site to 2022Q3, which does not appear affected. #status
We are working on a filesystem issue where adding disk space caused some member sites to become read-only because they think (incorrectly) that they have run out. #status
We will be doing some upgrades to our infrastructure supporting MariaDB. This will cause short disruptions for hosted MariaDB processes. We anticipate about 10 minutes for each process over the next two weeks. This will also help make future updates less disruptive. #status
We have deployed a filter designed to block the “ThemeGrill Demo Importer” vulnerability from affecting WordPress sites hosted on our service. But if you use or ever used this plugin, you should check your site and remove it. (Details: https://t.co/RogfGyUotZ ) #status
Also, during July & August we will rebasing the beta realms (white & black) to FreeBSD 11.3 so we will be ready for an 11.3-based 2019Q4 production realm.
We just got a phishing spam purporting to be from us containing the text “You have 15 delayed messages on 5/29/2019 5:52:43 a.m.. Your action is required,” with a link to click to “release” the supposed messages. If you get one too, it’s definitely a fake!
Relentlessly clever office favorite @somemorenews has a great look at the (totally predictable) negative effects of the SESTA/FOSTA attacks on free speech: https://t.co/VmKMjlPdtu
None of our PHP environments are affected by the go-pear.php PHP PEAR vulnerability announced today. We build everything from source obtained via the PHP git repository and use install-pear-nozlib.phar, which was not affected anyway.
Took the two affected machines offline to test; both run 35x slower than normal when on only one power feed, but CPUs report full speed. Identical machines in same rack are fine. That's a new one! Will replace PSUs and retest. (No service impact, just thought it was interesting.)
That is definitely not how redundant power works, but it's a heck of a coincidence! So we're trying to understand if/how there could possibly be a connection. In the meantime, we really appreciate our datacenter staff for their quick action to restore full power! <3
Our datacenter had a power failure this evening (~1:30-2:00am US EST). Our equipment is redundant; we did not lose power. But a couple front end machines were weirdly slow during the outage, causing intermittent timeouts for some people. #status
The problem should be resolved. Adding disks to storage cluster -> slow storage performance -> DNS timeouts -> database connection backlog -> outage.
A supposed-to-be routing upgrade of our core distributed filesystem did not work as advertised, causing widespread disruption to our service, but it is recovering and should be available shortly. #status
Quarterly realm rotation has occurred; the 2018Q2 stable realm is now available. This realm brings support for new languages to the stable track, including C#, D, F#, R, Rust and Scala. #status
FOSTA reviews are in. “It’s unconstitutional, ineffective, and does major damage to free speech online. 5 stars! We love it!” - The US House of Representatives. We disagree. See: https://t.co/ykyWYmHhxd