| |||||||||||||||||||||||||||||||
|
MegaMail By front-ending a coordinated, load-balanced mail server cluster, Bluetail Mail Robustifier virtually eliminates email reliability and scalability worries. It's green, but we think the benefits make it worth any ISP's while to take a look.
ISPs and their customers may not always see eye to eye, but on this point they agree: Email is a mission critical app. Woe be the ISP who takes mail server reliability lightly. Successful ISPs know that losing a mail server means losing customers. Most stand ready with a replacement server to limit downtime in the event of mail system failure. Bluetail Mail Robustifier goes one better: it allows a physical server cluster to work together as one robust, high-availability virtual mail server, insulating the ISP from single-server failure. High-Availability and Scalability by Proxy To understand how Bluetail Mail Robustifier works, consider this example configuration: a pair of front-end systems proxy mail on behalf of four back-end mail servers. Only the proxies are externally visible (i.e., associated with MX records). These proxies are configured into two clusters: an SMTP cluster with proxy "A" as primary, and a POP cluster with proxy "B" as primary. The four back-end servers participate in both clusters, allowing workload to be distributed across servers according to configured thresholds. If one proxy should fail or be taken out of service, the other will take over transparently, covering both SMTP and POP. Similarly, back-ends can be added or removed transparently, with proxies directing traffic to avoid server overload. This type of configuration offers redundancy and scalability when compared to stand-alone mail servers, and can be used with a combination of mail servers (e.g., sendmail on back-end "X", qmail on back-end "Y"). Load Balancing and Mail Session Routing If back-end mail servers share a common (e.g., NFS) message store, Bluetail balances load across all servers based on configured parameters like response time, run queue length, mail queue length, and disk utilization. If back-end mail servers use separate message stores, mail sessions must be routed to the server or servers who own a given mailbox. Mail sessions can be routed by user or domain, driven by mapping tables supplied in text format or drawn from an external (LDAP, SQL, or RADIUS) server database. If a back-end server goes off-line, proxies automatically re-route requests to other serversor reject requests if no suitable server is available. In fact, if Bluetail detects that a server is becoming overloaded (e.g., queue length or disk utilization at defined threshold), it re-routes or rejects traffic to avoid a server crash. This Mail Robustifier feature helps the cluster survive sudden bursts of legitimate or spam traffic that might overwhelm a stand-alone mail server. Different thresholds can be defined for groups of users, so that premium users can receive priority when regular users are being deflected during periods of heavy traffic. Custom Handling for Troublesome Domains &
Users With Bluetail, administrators can also configure separate SMTP mail queues for selected domains. When doing so, any mail previously queued to the affected domain is automatically moved to the newly-created queue, unclogging the existing queue so that others aren't affected by the troublesome domain. Bluetail Installation and Management All Mail Robustifier proxies run under the supervision of a management serverthe Bluetail Cluster Managerinstalled on one of the proxy systems. The management server provides both command line and Java-based graphical interfaces for single-point configuration and monitoring. Configuration changes and even software updates can be completed without shutting down the mail system. Changes are entered at the management server and then internally propagated to cluster memberseven those that happen to be down at the time the change is made. All proxies record statistics, events, and alarms in a local log. Local logs merged and stored by the management server, providing a consolidated record and view of mail system behavior. The newest version, Bluetail, Mail Robustifier 2.1, also supports SNMP traps for admin notification of exceptional events and audit trail logging to track configuration changes. Pricing and Support Bottom Line End |
|
|||||||||||||||||||||||||||||
|
|
|||||||||||||||||||||||||||||||