An old SATA SSD started throwing read errors. Looking back, the warning signs had been in the SMART log for weeks.
sudo smartctl -a /dev/sda | grep -Ei 'reallocated|pending|wear|crc'The attributes worth watching are reallocated sectors, pending sectors and the wear indicator. A rising CRC error count usually means a bad cable rather than a bad disk. Now a weekly job mails me the short summary.