[ https://issues.jboss.org/browse/ISPN-2442?page=com.atlassian.jira.plugin.... ] Dan Berindei commented on ISPN-2442: ------------------------------------ {{SITE_UNREACHABLE}} messages are still ignored for async backup sites, as the backup take-offline policy can only act on individual RPC failures and there is no way to connect a {{SITE_UNREACHABLE}} message to a previous async backup command. As a compromise, we could make {{SITE_UNREACHABLE}} trigger a failure on all the async backups to that site. To keep the take-offline policy settings independent of the number of caches, we could assign the failure randomly to one of the async backups to that site (or to none of them if one of the sync backups has outstanding requests). It would be even better to define the site and the take-offline policy at the cache manager level. That way all the {{SITE_UNREACHABLE}} messages will be counted as failures against the same target, and the backup site will be taken offline at the same time for all the caches. -- This message was sent by Atlassian JIRA (v7.5.0#75005)