Post #3635106

@AndresFreundTec@mastodon.social @gregkh@social.kernel.org @joshbressers@infosec.exchange after meltdown/spectre we implemented Debian unattended upgrades in our infrastructure (1500+ servers) with a hook system around it so that servers can drain and undrain themselves. We made it to 75-80% system coverage over time. We always do a reboot, no matter what’s actually installed in the UU run. That took a lot of pressure from the whole story. Before that, evaluating CVEs and their impact was a daily job.