We are regularly enhancing the IBM ECDN service with new capabilities and fixing defects. Here are the new changes across all three components of ECDN:
- ECDN Management Portal – Portal
- Virtual appliance for the ECDN server – Server
- ECDN backend services – Backend
Server ECDN now supports SSH-key based logins into servers. With this ECDN servers will be more compliant with enterprise security standards.
Portal In line with customers’ needs we have added a brand new user management system. Now within ECDN portal three different roles per user can be added: SuperAdmin, Admin and Reader. Additionally, all SuperAdmins can now see and manage the list of all ECDN users. You can learn more about user roles in the support article at http://ibm.biz/ecdn-user-roles
Portal You can now define a recurring daily or weekly maintenance window in the ECDN portal. At the specified time, an automated process will be initiated to download all available upgrades and security patches. This will help ensure all ECDN servers in your network are running the latest release, and have all known security patches installed. We also allow you to temporarily disable automated upgrades to meet your enterprise network freeze requirements, typically seen during end of quarter or end of year.
Server Some enterprise IT guidelines enforce strict naming rules for ECDN servers. In these cases, you will have to yourself manage the custom domain name. You will have to additionally order a SSL certificate for this domain, and manually install them on ECDN servers. Read more about how to setup SSL certs on ECDN servers.
Portal Manually control what ECDN servers, a player can connect to. You can now specify a whitelist of blacklist of private IP addresses as part of an ECDN location definition. Once enabled, ECDN servers will only respond to playback requests as per these rules. Read more about how to setup these rules.
Server ECDN service is now GDPR compliant.
Portal Added support for historical usage reports. You will now be able to retrieve ECDN usage reports for any duration within the past two years. Read more about Getting historical usage reports.
Portal You can now optionally enable using a disk based cache for storing the recorded videos. The cache contents will not be lost when the ECDN server is rebooted. Previously, with memory based cache, VOD cache contents were lost when the server was rebooted. VOD contents are cached on first request. When full, the oldest content is deleted to make space for the new video being streamed. To enable this feature, just add a second hard disk to the ECDN server VM in the hypervisor and restart it. No other configuration steps are needed. Read more about enabling VOD cache persistence to disk.
Portal Server list page displays which ECDN servers have enabled the VOD cache persistence to disk. Server details page displays the size of the cache disk, how much of it has been used.
Server New version to fix the Meltdown and Spectre vulnerabilities.
Backend In case of CDN fallback, ECDN servers will use the new IBM Watson Media content delivery network (CDN) that is available worldwide.
Portal When configuring a new ECDN location, an IBM Watson Media midgress region should be selected by the customer from now on. For existing locations, a default value has been defined for midgress region.
Server We have added support for midgress regions in the virtual appliance.
Portal Support for "multi-tier topologies" of ECDN servers. Now you can organize the locations of ECDN server instances in a multi-tier hierarchy. This gives you the flexibility in defining the video content distribution architecture that matches your company's unique network design. ECDN servers can be configured to a maximum of 3 levels (two levels of mid-tier servers + 1 leaf level). Read more about multi-tier deployments of ECDN servers and how to set them up in a multi-tier hierarchy.
Backend The IBM managed domain name has changed to
deepcaching.net. For all customers who pick the option for IBM managed domain, a domain name like
*.<customerName>.deepcaching.net will be assigned to their ECDN server instances.
Portal New "Update remotely" option added for propagating configuration changes from portal to the edge server instance. With this option the edge server auto-detects the configuration changes, downloads and installs it. The manual option, while still available, is no longer needed where you had to manually download the virtual CD image file, then attach it to the edge instance VM in the hypervisor, and reboot the VM to complete the change. The web portal displays "Pending changes" status for the edge server, while the changes are being propagated. Learn more about configuration update options for ECDN servers.
Portal All new configuration setting values are highlighted in a different color. The highlights are removed when the changes are successfully propagated to the edge server instance.
Server Edge servers now have remote configuration update capabilities for propagation of configuration changes.
Portal New "Remote assistance" checkbox. You now have control to enable/disable the VPN tunnel connection to ECDN backend for each edge server instance. By enabling Remote assistance, you can let ECDN support team remotely login to the edge server instance, and help with debugging and fixing it.
Server Edge servers have remote assistance capability to control the VPN tunnel connection.
Portal The status for all service daemons is now displayed in the edge server details page. New additions to the list of services are: Metrics service, SNMP polling, Log forwarding and Remote assistance.
Portal The "System health" widget that indicates the overall health of the edge server instance, now shows "Poor" status when the edge server instance is under load. It indicates that the edge server may be at capacity, and may need additional hardware resources. This aggregate status is based on: available memory, available disk space used by caching service, and number of player connections.
Portal In the "Edge servers" widget, we now display the health status details popup when your cursor hovers over the aggregate health status dot indicator in the "Health" column. This makes it easier to identify the unhealthy services. Now you don't have to go to navigate to the edge details page to get this information.
Server Now, by default, all edge server instances automatically failover to IBM Watson Media content origin servers, whenever there is an outage in any external 3rd party CDN vendor service.
Server Each edge server now uploads health status of various service daemons running inside it to the ECDN backend services.
Server Improved memory management of edge servers ensures that more memory is available to cache the video content.
Server Edge server instances now cache the DNS responses, thereby reducing the calls to customer's DNS servers. This makes the ECDN edge service more resilient to DNS outages.
Backend When no heartbeats are received for more than 2 minutes from an edge server, from now on, that instance is marked as unavailable to serve video traffic.
Backend Fixed the issue where ECDN portal would not show any metrics in the Player and buffering graph for edge servers that had any upper case letters in their hostnames.