Resolver v3.2.0 + <70GB disk: blocking risk (action required)

We’re following up on the previously communicated issue affecting local DNS resolvers upgraded to 3.2.0 on hosts below the existing ≥70GB disk requirement.

On such hosts, components responsible for DNS blocking may enter a restart loop after the upgrade. As a result, DNS resolution remains available, but DNS blocking may be partially or fully disrupted, including cases where newly added malicious domains are not blocked. You may also observe higher CPU usage. Hardware sizing recommendations documentation.

What we changed (prevention)
To prevent additional impact, we have introduced an Admin Portal safeguard that blocks upgrades on resolver hosts that do not meet the ≥70GB disk requirement (based on resolver sysinfo). This is a temporary protection and will be complemented by an resolver agent-side check later.

What you need to do (if already upgraded / impacted)
Immediate hot-fix (recommended):

• In Admin Portal → Resolver → Advanced configuration → Expert settings set: LOGROTATE_SIZE=400M
• Apply to all resolvers and restart/roll the resolver if needed.

Long-term fix: ensure the resolver host meets the ≥70GB disk requirement for stability and future upgrades.

If you need help identifying impacted resolvers or applying the configuration, please contact Support.

WARNING: 3.2.0 may fail on hosts below the existing ≥70GB disk requirement

Starting with 3.2.0, resolvers include a new feature that automatically sizes the local threat database (LMDB) to allow a wider database / better coverage. To keep this safe, the resolver now validates available disk space during startup.
On hosts that do not meet the minimum requirement (≥ 70 GB disk), particular components may restart-loop and DNS blocking can partially or fully stop (including newly added malicious domains not being blocked) and you may see higher CPU usage.

Required action

• Ensure the resolver host has ≥ 70 GB disk, or
• Set LOGROTATE_SIZE to 500M in Admin Portal → Resolver → Advanced configuration, apply to affected resolver(s), and restart/roll if needed.