zSeries features for optimized sockets-based messaging: HiperSockets and OSA-Express

Abstract

In recent years the capacity of mainframeclass servers has grown, and the quantity of data they are required to handle has grown with them. As a result, the existing S/390 ® I/O architecture required modifications to support an order of magnitude increase in the bandwidth. In addition, new Internet applications increased the demand for improved latency. Adapters were needed to support more users and a larger number of connections to consolidate the external network interfaces. The combination of all of the above requirements presented a unique challenge to server I/O subsystems. With the introduction of the zSeries TM comes an enhanced version of a new I/O architecture for the mainframe called queued direct I/O (QDIO). The architecture was initially exploited for Gigabit and Fast Ethernet adapters. More recently the architecture was exploited by the OSA-Express network adapter for Asynchronous Transfer Mode (ATM) and highspeed Token Ring connections, and it was exploited by HiperSockets for internal LPARto-LPAR connections. In each of these features, the TCP/IP stack is changed to tightly integrate the new I/O interface and to offload key TCP/IP functions to hardware facilities. For external communications, the offloaded functions are performed by the OSA-Express hardware microcode; for internal communications, the offloaded functions are performed in the zSeries Licensed Internal Code (LIC). The result is a significant improvement in both latency and bandwidth for sockets-based messaging which is transparent to the exploiting applications.

Libra

“…All three represent microcode-implemented optimizations for hardware. In the case of IFL, the Linux systems run on logical partitions on top of the specialized processors, and use HiperSockets [6] to communicate with each other as well as other processors on the mainframe.…”

IBM System z9 Open Systems Adapter for Communication Controller for Linux

“…The solution must remain transparent to the host operating systems and NCP, where each host must continue to use the existing channel protocol called channel data link control (CDLC). ; its main function is to bridge an operating system image using CDLC protocol to the CCL in the Linux image using the queued direct I/O (QDIO) architecture [8]. The OSN channel provides the CDLC channel with connectivity between logical partitions (LPARs) or virtual machines in the same mainframe system.…”

“…2. QDIO Y side: Here, the CCL image, using the QDIO architecture [8], communicates with the OSN, which connects to the host OS. The system device support appears as an OSN device.…”

“…For a 3745 write operation, OSN has transferred all of the SNA PIUs (path information units, which describe SNA messages) for the entire write channel program into OSN memory from the host OS via the 3745 device. OSN then repackages the data (adding the QDIO header) and transfers it to the Linux image using the OSN data device via QDIO queue structures [8].…”

FCP for the IBM eServer zSeries systems: Access to distributed storage

“…An overview of the QDIO protocol is given in [12]. However, since QDIO was initially defined for networking traffic, it did not fully meet all of the needs of storage attachments and had to be extended accordingly.…”