Hello,

We run ClamAV in AWS Lambda (ca-central-1) scanning document uploads for a
Canadian healthcare product. Since ~Aug 19, freshclam from our Lambda
egress receives Database update process failed: Forbidden; Blocked by CDN
with a 24-hour cool-down. The fault was originally ours: no persistent
datadir meant a full database download per cold start, retried on a 6-hour
schedule. We have fixed that — the updater now seeds its datadir from our
own storage so fetches are incremental cdiffs, persists the cool-down and
honours it *before* any contact, and attempts at most once daily. The block
has outlasted the fix; the same freshclam build works immediately from
residential and cellular networks.

In the interim we mirror definitions from the official clamav/clamav Docker
image — but we note its bundled databases can lag: all current tags shipped
daily 28101 (Aug 23) as late as Aug 29, while daily 28107 was current.

Three questions:

   1. Is there a path to restoring CDN access for a shared-egress (AWS
   Lambda NAT pool) environment once client behaviour is corrected, or is that
   address space effectively unusable for freshclam?
   2. Do you allowlist specific static egress IPs? We can place the updater
   behind a fixed NAT address if that is the supported route.
   3. What is the recommended pattern for serverless ClamAV deployments
   generally? Ephemeral compute plus shared egress seems structurally awkward
   against per-IP controls, and we would rather follow a supported design than
   improvise one.

Happy to provide timestamps/details off-list. Thanks — Paul Theron
_______________________________________________

Manage your clamav-users mailing list subscription / unsubscribe:
https://lists.clamav.net/mailman/listinfo/clamav-users


Help us build a comprehensive ClamAV guide:
https://github.com/Cisco-Talos/clamav-documentation

https://docs.clamav.net/#mailing-lists-and-chat

Reply via email to