195295fm.exe   Server 195/295 Field Maint. Test 2.0
295diag.exe       Server 195 and Server 295 diagnostic diskette
295fw170.exe   Server 295 FirmWare upgrade version 1.7
295ref.exe         Server 195 and Server 295 reference diskette version 1.7.1
295start.exe      Server 195 and Server 295 system startup diskette
mpext21f.exe     Server 295 HPFS fix for MP extensions
mass160.exe      Server 195 and Server 295 MASS/2 version 1.60
raidnw.exe         Server 195 and Server 295 RAID installation diskette for NetWare

750 watt universal power supply;  Opt redundant power supply.

NOTE: The PSU is Universal, but NOT Auto-Ranging. There is a set of spade terminals on the PSU's PCB, connected for 115v, open for 230v.

Processor:
8600-001 – 486DX-33 w/128KB L2 , 486DX-50, w/256KB L2.
8600-002 – Two 486DX-50s w/256KB L2 cache. (L2 direct mapped, write-through)
ASMP for OS/2 (one 486 for OS and apps; one 486 for file system and protocols).

Each processor board has a switch on the mounting bracket, AP or FP.
AP - Application Processor. Runs OS/2 programs
FP - File Processor. Performs the file related functions.

Each processor board contains a 486DX processor, along with a Level 2 memory cache, and the attachment to the Micro Channel. There are two Micro Channels, one with four slots, and the other with eight slots. The processor in slot 0 is always connected to the eight-slot Micro Channel, and the processor in slot 1 is always connected to the four-slot Micro Channel.

Configuring the Server 295 for multiprocessor or uniprocessor mode is done using the utility MPSETUP, which is also used for installing Multi Processing Extensions/2 and for allocating memory to the AP and FP processors, and the HPFS cache. 

When configured for multi-processor mode, changes are made to the CONFIG.SYS file, and a new CONFIG.FP file is created for the FP processor. However, the utility MPSETUP should always be used to switch between uni-processor and multi-processor modes. It is wise to always return to uni-processor mode before altering the CONFIG.SYS file or upgrading the operating system.

ECC-P Memory with 4MB 30 Pin SIMMs
The memory card has (32) 30 pin SIMM sockets. Each 32 MB (two banks) has own memory controller. Memory was sold in 32MB increments. That would be two blocks with four 30 pin SIMMs.

32+4=36, not enough for ECC, BUT... 36+36=72... 64 bits for data, and 8 bits for ECC... Two 32 bit EDCs can be chained for a 64 bit word.

So... you need two banks of 4 SIMMs each for ECC-P, so the minimum is 32MB, the next is four banks (64MB), six banks (96MB), and 8 banks for a maximum of 128MB.

NOTE: I have seen a configuration screen, with a choice of "ECC: ON | OFF". My guess is that ECC=P will not run on an odd number of banks, so if you wanted to run 48MB, ECC would be "OFF".

Previous IBM servers such as the IBM Server 85 were able to use standard memory to implement what is known as ECC-P.  ECC-P takes advantage of the fact that a 64-bit word needs 8 bits of parity in order to detect single-bit errors (one bit/byte of data). Since it is also possible to use an ECC algorithm on 64 bits of data with 8 check bits, IBM designed a memory controller which implements the ECC algorithm using the standard memory SIMMs.

The following shows the implementation of ECC-P. When ECC-P is enabled via the reference diskette, the controller reads/writes two 32-bit words and 8 bits of check information to standard parity memory. Since 8 check bits are available on a 64-bit word, the system is able to correct single-bit errors and detect double-bit errors just like ECC memory.



While ECC-P uses standard non-expensive memory, it needs a specific memory controller that is able to read/write the two memory blocks and check and generate the check bits.  Also, the additional logic necessary to implement the ECC circuitry make it slightly slower than true ECC memory. Since the price difference between a standard memory SIMM and an ECC
SIMM has narrowed, IBM no longer implements ECC-P.

Memory Management and OS/2 version 1.3
Up to 128MB of 80ns ECC-P memory can be installed in the Server 295.  Although OS/2 1.3 can only address up to 16MB of this memory per processor, the rest can be accessed via the Transparent Reserved Memory Manager (RMM) utility.

When configured in multiprocessor mode, up to 16MB of memory can be assigned to each of the FP and the AP processors, for a total of 32MB.  This allocation is done using the utility MPSETUP.

The remaining memory can be accessed via RMM, which is installed as a device driver.  RMM enables this extra memory to be accessed in a number of ways:

o Disk buffers for the PDAs
o HPFS cache
o Reserved memory for some OS/2 applications - for example in response to DosAllocSeg and DosAllocHuge segment requests
o RAM-based virtual disk
o SWAPPER.DAT file used for segment swapping to disk by OS/2

RMM memory is allocated by the /TRMM= switch in the CONFIG.SYS and CONFIG.FP files.  Status information on RMM can be displayed using the RMMSTAT utility.

RMM is needed because of the 16-bit limitations of OS/2 1.3, and results in substantial performance improvements for some applications because massive amounts of data can be held in memory rather than on disk.

Bus Architecture:
64 bit 200 MB/sec Interprocessor Bus; hierarchical design to interconnect the processor(s), memory, SCSI controllers, and Remote Maintenance Processor (RMP); 6 slots.
Parity protection on data, address, and control buses (IP-Bus).
Each 486 has independent 20 MB/sec Micro Channel bus (8 slots on 1st 486 + 4 slots on opt 2nd 486 = 12 total slots).

Disk/Controllers/Bays:
 Two independent (dual channel) RISC-based SCSI I/II controllers on single card; 64 bit interface to IP-Bus.

NOTE: I see two NCR 53C700 SCSI controllers. Not enough clarity to see if there is a RISC controller or not.

Option: 2 additional SCSI controllers (same as above, on 1 card).

Supports IBM 400 MB (11.5 ms) and 1 GB (11ms) SCSI disks.

Supports 1 to 28 hot insertion/extraction disks (all hot pluggable).

Max of 9 GB internally (9 bays x 1 GB disk).

Max of 28 GB total (4 SCSI x 7 disks (1GB)) with Exp Cabinets.

Ten 3.5" half height bays internally (diskette uses one).

Supports 1 to 3 optional External Expansion Cabinets each with ten 5.25" half height or five 5.25" full height bays.

OS/2 preload includes FTUTIL: allows disk pairs to be striped, mirrored, or duplexed. Also allows hot spare pooling, hot insertion/extraction, automatic data rebuild, and hot fix.

The disks in the PS/2 Servers 195/295 can either be used as individual disks, configured as hot-standby spare disks, configured as Parallel Disk Array (PDA) pairs using the FTAUTIL.EXE standard utility, or combined into RAID-5 PDAs using the optional Orthogonal RAID-5 Disk Array/2 software, which implements orthogonal RAID-5 fault tolerance. Data and parity information are spread across the disks in a PDA, thus both protecting the data and providing performance improvements for applications which mainly read data (rather than write data).  The PDA can be set up to be "orthogonal" by ensuring that the disks of the array are on separate disk controllers, thus increasing the availability of the system still further.

One or two disk controllers can be installed on the IP-Bus, providing two or four SCSI disk channels, respectively. Each channel can support up to seven SCSI disks, for a maximum of 28 disk drives.

PDAs are set up using the Server 295 reference diskette.  Information on the PDAs can be displayed using MASS/2.  Configuring spare drives and reconstructing a PDA following a drive failure is done using the utility PDAUTIL, as well as automatically by MASS/2, if there is a hot-standby spare disk in the system.

Orthogonal RAID-5


Orthogonal RAID-5

Orthogonal RAID-5 is an enhancement on RAID-5, which has been implemented in the PS/2 Server 195/295 internal disk array.

The performance of a disk subsystem depends on more than just the underlaying performance of the disks. Multiple requests to one disk or across one adapter will typically take longer to satisfy than the same number of requests to multiple disks across multiple adapters.

In addition, to overall reliability of a standard RAID-5 system is dependent on the reliability of the one disk adapter to which all of the disks are connected.

Orthogonal RAID-5 solves both of these concerns by grouping the disk arrays orthogonally  to the disk adapters. SCSI buses and power cables. This would best implemented as a four-drive orthogonal RAID-5 array, where each disk would be connected via a different adapter and SCSI bus.

The result of this is that any one component of the disk subsystems, not just a disk drive, can fail with no loss of data and no interruption to system operation.

RAID Performance Characteristics

RAID Level	Capacity	Large Transfers	High I/O Rate	Data Avail.
Single Disk	Fixed (100%)	Good	Good	Note 1
RAID-0	Excellent	Very Good	Very Good	Poor Note 2
RAID-1	Moderate (50%)	Good	Good	Good
RAID-2	Very Good	Good	Poor	Good
RAID-3	Very Good	Very Good	Poor	Good
RAID-4	Very Good	Very Good	Poor	Good
RAID-5	Very Good	Very Good	Good	Good
Orthogonal RAID-5	Very Good	Very Good	Good	Very Good

MASS/2 Features (Maximum Availability and Support System/2)

Integrated software allowing monitoring, controlling, tuning, and recovery of 295. Allows 295 to be geographically distributed while permitting comprehensive, centralized control with mainframe-like remote administration. Features:
 Failure recovery without human intervention (most cases).
 Operation even when 295 powered down (RMP with battery).
 Concurrent access by multiple users.
 Access via console, across LAN (local), via modem (remote).

  Continual monitoring of CPU, memory, disk, network, and swapping level utilization via realtime bar graph, and 1 hour history in RMP RAM, and historic logs on disk.

  Continual monitoring of temp, PSU voltage, single bit memory errors, network errors.

  Comprehensive configuration details (processors, memory, SCSI, arrays, spares, RMP, UPS status, adapters).

  Alarm and threshold selection with automatic dial out capability for various events like disk failure, disk full, reboots, shut downs, memory errors, over temperature, power low/high, and network errors.

  Scheduled power down and power up; remote rebooting.

  SAA-compliant graphical user interface (hierarchical).

  Access to network operating system logs and server logs.

  Multilevel password based security; G File transfer capability.

 Preinstalled on Server 295.

Inside the Parallan Super Server

"Superservers: Finding a Home in User Networks" Network World Jun 3, 1991 Page 23

Parallan Server 290 Architecture

 "Super Servers Waiting in the Wings",  InfoWorld Apr 6, 1992 Page 48