• Status: Solved
  • Priority: Medium
  • Security: Public
  • Views: 855
  • Last Modified:

IBM SCSI drive problem on NW312 server

I've recently added an IBM DNES-309170 9GB wide SCSI drive to our NW312 server, which previously had two Seagate ST15150N narrow Scsi drives (these are still attached).
These two ran flawlessly for years with the SYS: volume shared between them. Now the SYS volume is spanning all 3 drives.

The Seagates share a 50 pin narrow scsi cable (last with termination on), the new drive is second-to-last device on the 68 pin cable, with active terminator at end of cable - the IBM drive has no internal terminator option. The Controller card is an Adaptec AHA2940UW.

When there is normal disk activity the server runs fine, but at night when disk usage drops to near zero, the SYS: volume dismounts with a message: DRIVER DEACTIVATED DUE TO DEVICE FAILURE: IBM DNS309170.

The disk utilities built into the SCSI card BIOS check the disk drives and find no flaws at all.

I've tried various options in the Adapted SCSI BIOS setup:
Disconnect: on and off,
Transfer rates from 5 to 40 MB/s,
Sync negotiation: on and off,
Wide transfer initiation: on and off
Tag queueing disabled and enabled

I tried the wide SCSI IBM disk shared on the narrow SCSI cable, with 68-50 pin adapter; tried a different SCSI controller card; & the IBM drive is set to Single ended mode, not LVD.
Nothing seems to have any effect... this problem ONLY occurs a few hours after activity on the SYS: volume drops to near zero.. if I leave a backup running overnight - no problem!

The only thing left to try is TERM PWR ENABLED on the IBM drive, but Adaptec say with the AHA 2940 this is not necessary.
Any suggestions PLEASE!?
1 Solution
davejhbAuthor Commented:
Oh ps: The drivers for the SCSI card are Adaptec's latest, but using old ones didn't  make any difference either.
About two years ago I ran into the same problem with a rather large Seagate drive. I went through the usual Adaptec support and Seagate support meanwhile getting more and more confused. The client's server ran flawlessly during the day but at night, the Seagate disconnected. We also use the Adaptec 2940. Try going into the Adaptec setup at boot and making sure that drive translation is turned off for that drive. This worked for us. You will have to re-partition the drive. Good Luck
Try to read this TID's ,if they can  help you.


Tips for Adaptec drivers (Last modified: 16AUG1999)  TID2945634
This document (2945634) is provided subject to the disclaimer at the end of this document.

This document gives some recommendations on improving reliability and performance of NetWare servers using Adaptec AIC78xx based SCSI controllers. This includes most Adaptec AHA29xx and AHA39xx adapters with the exception of the AHA2920A. This information has been gathered from a variety of sources and has not been officially verified by Adaptec or Novell.


There are 2 different families of AIC78xx based controllers, using different drivers:

1/ All Adaptec Ultra2 controllers. For these controllers, the only available driver is AIC78U2.HAM. See the readme file in 7800NW.EXE for a complete list.

2/ All the other AIC78xx based controllers. For these controllers you can choose between the AIC7870.DSK driver or the AHA2940.HAM driver. At the time of writing the latest drivers were:

On Adaptec's web site ( http://www.adaptec.com ), in the file 7800NW.EXE:
          AIC78U2.HAM Version: 6.10  (980911) and AIC7870.DSK Version: 4.20  (980902)

On http://support.novell.com/, Minimum Patch List, locate the file IWSP6a.EXE. This file contains:
          AIC78U2.HAM Version: 6.00  (980622) and AHA2940.HAM Version: 7.00  (980720)

For those who do not wish to apply the whole Support Pack they are also in nwpaup1a.exe; use the File Finder on the above website to search for it by name.

Important Notes:

1) Always use the most current drivers, i.e., AIC7870  (1998) or the .HAM drivers.

2) Some computer manufacturers that integrate Adaptec SCSI controllers on their motherboards have customized their SCSI BIOS to the extent that the controller no longer works with Adaptec drivers. In this case, a customized disk driver has to be used. These customized disk drivers are recognized by an "S" in their version number (eg: AIC7870.DSK v2.13S9). In this case you have to refer to the computer manufacturer for updated drivers.

HAM or DSK drivers ?

For Adaptec Ultra2 SCSI controllers, HAM drivers are the only choice. For the other controllers you have the choice between the AIC7870.DSK driver and the AHA2940.HAM driver (note that with NetWare 5 only HAM drivers are supported). There is no general recommendation on which one is better, but here are some ideas:

- If you decide to use HAM drivers, you should install nwpaup1a.exe or IWSP6a.EXE, with the disk and lan driver update option checked

- In some cases HAM drivers may be faster

- Some third party backup software does not officially support HAM drivers for Adaptec controllers. Some people have reported that these programs work correctly with the nwpaup1a.exe/IWSP6a update. Without the updates these programs may not work reliably.

- HAM drivers are IRQ 15 safe.

SCSI CDROM support

When using HAM drivers, SCSI CDROM support is automatic through the SCSICD.CDM driver provided by Novell. If you are using the AIC7870.DSK driver, however, you may have to enable CDROM support yourself. With older versions of AIC7870.DSK, this was done by loading the ASPICD.DSK driver. Since version 4.0 of AIC7870.DSK, CDROM support is built into the AIC7870.DSK driver. You can enable it by adding the option "cdrom=on" to the line loading the AIC7870.DSK driver (DEFAULT SETTING IS CDROM=ON).

Other important command line options

This option is mandatory for the HAM drivers. In addition, you have to load the HAM driver once for each SCSI controller present in your server, each time with a different SLOT= parameter. For the AIC7870.DSK driver v4.10 or later, the SLOT= parameter is optional. If you specify the SLOT= parameter, you have to load the driver once for each controller in your system. If you omit the SLOT= parameter, loading the driver a single time will handle all controllers supported by this driver.

By using this option, you disable tagged command queuing, an optimization technique which is supposed to improve performance. See TIDs 500362 and 2925042 for more information on issues related to this parameter.

This option is for .dsk drivers only, it will cause the .ham drivers to fail to load if used.

Use this option if you have a device that needs LUN addressing, such as SCSI CDROM changers and external RAID controllers.

CDROM=ON  (AIC7870.DSK only)
Use this option to enable support for SCSI CDROM drives. In this case do not load ASPICD.DSK

Using this option will disable read after write verification at the controller level. All modern SCSI drives do read after write verfication at the drive level, and doing it twice results in reduced performance with no gain in data reliability. This performance difference is especially noticeable in mirroring or duplexing configurations; in that situation this option is a must.

If either the controller or the drive are performing Read After Write Verification then the NetWare Set parameter that implements RAWV functionality in the OS should also be disabled; use Servman (NW4.10) or Monitor (NW4.11 or NW5 only) to do this.

For a complete list of command line options, check the file README.TXT included in 7800NW.EXE

Related TIDs:

HP NetServer 166 LS Page Faults with AIC7870 - TID2919876
HP Server Hang During Remirror - TID2934806
AIC7870 volume disappearing/dismounting - TID2935457
Latest Adaptec 7700&7800 Scsi drivers(8/97) - TID2924897
Servers Hang After IWSP4a Upgrade - TID2936714
LOADER Abends With Support Pack 4 - TID2935517
AUTOEXEC.NCF Not Executing from SYS:SYSTEM - TID2935907
Slow Core Dumps on HP LX Server - TID2935460
Loader cannot find public symbol ASPI_Entry - TID2934553
NetWare Support Pack v6.0 - TID2944744
NWPA.NLM and CDROM.NLM updates - TID2942518
Support Tips For Adaptec EISA, VL, PCI HBAs - TID500362
Adaptec 2940 / Quantum Drives - TID2927805
HP Netserver LH and AIC7870.DSK - scsi errors - TID2925018
EMC Disk Farm parameters - TID2930913
AHA-2940 Hardware Setup Tips - TID1001599
Adaptec 1540 hangs server - TID2928579
ADAPTEC 2940 - TID2919935
ADAPTEC 7870 - TID2926455
Mirroring and Performance - TID2924724
No CDROM Devices Registered With Netware - TID2921568
HP Netserver Unable to load AIC7870 and HPDA4 - TID2923850
ADAPTEC 2940W redirected blocks - TID2913106
Adaptec and NetWare v3.11 Installation Issues - TID2925042
What is Device Deactivation - TID500051

Document Title:      Tips for Adaptec drivers
Document ID:      2945634
Document Revision:      6
Creation Date:      06DEC1998
Modified Date:      16AUG1999
Novell Product Class:      NetWare

Novell Product and Version:      NetWare 3.12
      NetWare 3.2
      NetWare 4.1
      NetWare 4.11
      NetWare 4.2
      NetWare 5
      NetWare for Small Business 4.11
      NetWare for Small Business 4.2
      intraNetWare 4.11


Support Tips For Adaptec EISA, VL, PCI HBAs (Last modified: 26AUG1998)  TID500362
This document (500362) is provided subject to the disclaimer at the end of this document.
Abends, server hangs, corrupt data, corrupt file system, missing volume segments and drive deactivation are common symptoms experienced by users.
1.  From the NETWARE area on Adaptec's BBS download NW31X.EXE or NW40X.EXE to use the latest applicable driver.

      AIC7770.DSK      AHA-274x (EISA HBA), AHA-284x (VL HBA)  series
                                                and AIC7770 embedded chip set

      AIC7870.DSK      AHA-2940 (PCI HBA)

2.  Use the "tag_disable=ffffffff" switch for the AHA274x, AHA284x, and AHA294x, AHA274x T (T is for twin channel) or any "WIDE" channel HBA.  The surest way to disable tagged queuing is to use "ffffffff" (8 f's) for all drivers except the AHA1740.DSK which would require "tag_disable=ff".   Novell Labs has noted that some SCSI drives report to the host adapter that they support tagged queuing, however they fail to operate properly.  While the fault is not with the HBA and driver, it requires disabling tagged queuing. (Example "load aic7870.dsk tag_disable=ffffffff").

3. Set the Maximum Sync Transfer Rate to 8.0 MB/sec for the AHA-2940.

The AHA-2940 operates at greater than 10.0 MB/sec when configured at 10.0 MB/sec.  Since the SCSI spec only requires devices to operate at 10MB/sec, problems may arise.  With its BIOS enabled, the AHA-2940 will present a banner at system boot up time:  
<<<Press <Ctrl> <A> for SCSISelect  (TM) Utility!>>>
Start the SCSISelect utility, choose Configure/View Host Adapter Settings then under Additional Options, choose SCSI Device Configuration.  From the SCSI Device Configuration menu change the Maximum Sync Transfer Rate from 10.0 to 8.0 for devices #0 through #7.
4.  From the HARDWARE area on Adaptec's BBS download ASWC274.EXE.  This file contains the latest EISA configuration files for the AHA274x series HBA.  There are two EISA configuration issues which should be addressed.

     4.1  These configuration files will set the BCLKS=44 as the default.  Use the BCLKS=44 rather than BCLKS=60.  This is particularly important if the host machine is supporting multiple adapters.

     4.2  These configuration files support EDGE triggered interrupts.  Run the EISA config utility and set the AHA274x  HBA interrupts to EDGE triggered.

5. Avoid INT 15.  Both Novell and Adaptec have recommended avoiding INT 15 when problems arise.  Some users have successfully used INT 15, however, others have had problems.  Any time a user is having problems with an adapter using INT 15, AHA2740 or otherwise, recommend that the adapter be used at another interrupt.

6. Do not share interrupts.  The AHA-174X and AHA-274X may have difficulty when multiple adapters are installed in the computer and they are EISA configured to share interrupts.  If two adapters are installed they will be sharing interrupts by default.  The user must run the EISA configuration utility to change the interrupt on one of the adapters.

Document Title:      Support Tips For Adaptec EISA, VL, PCI HBAs
Document ID:      500362
Document Revision:      5
Creation Date:      08NOV1994
Modified Date:      26AUG1998
Novell Product Class:      NetWare

Novell Product and Version:      NetWare 3.11
      NetWare 3.12
      NetWare 3.2
      NetWare 4.1
      NetWare 4.11
      NetWare 4.2
      intraNetWare 4.11

Drive Deactivation Troubleshooting Tips (Last modified: 20FEB1998)  TID500089
This document (500089) is provided subject to the disclaimer at the end of this document.
Drive Deactivation Tips and Techniques for NetWare 3 & 4.

The purpose of this document is to establish what drive deactivation is and why it occurs.  Drive deactivation is when one or more devices (drives) in the server become unable to communicate with the operating system (OS). When the driver cannot communicate with the drive, it will normally retry several times.  If, after a predetermined number of retries, the driver still cannot establish a communication link with the drive, the driver tells the OS that the drive is not responding and the OS then deactivates the device.

Drive deactivation can occur across any of the disk interfaces, such as IDE, ESDI, or SCSI.  Drive deactivation is also not confined to just one manufacturer.  The following list is a "checklist" that gives suggestions to go through in troubleshooting a drive deactivation problem.  In many cases, specifics are required in regards to hardware, firmware, and software revision levels as third-party Host Bus Adapters (HBAs) manufactures and drive manufacturers are used in solving problems.  The priority of the checklist is based on solutions that will cost the least and still receive the highest priority.  For checklist purposes, this discussion will be restricted to the SCSI interface.  All suggestions should be easily translated for any interface.


1. Verify that termination along the SCSI bus is correct.  The first basic rule to remember is that both ends of the bus must be terminated.  Be aware that if there are drives in an internal or external cabinet, termination can be removed from the individual drives and then just terminate the cabinet externally.  This avoids future headaches because, as more drives are added, there won't be any confusion about which drive was terminated in which cabinet.

Another problem arises when both the internal and external ports on an HBA are used.  When this is done, the termination must be removed from the HBA and then terminate both ends of the SCSI bus, which will usually be at the drives themselves unless an external cabinet is used.

2. Verify that the SCSI IDs are set properly.  In all situations, the HBA will have a SCSI ID of 7.  This may be set either through a hardware setting (jumpers) or through software utilities (such as EISA CONFIGURATION or REFERENCE).

In PS/2 situations and possibly others, the HBA takes a SCSI ID of 7 and the attached devices start at 6 and work down to 0.  In all other situations, the HBA takes an ID of 7, and the attached devices start at 0 and work up to 6.

In the PS/2 environment, there are specific things that need to be addressed.

 - Under reference, the "fairness" needs to be turned ON for the HBA.

 - Using the OPTION diskette v1.0 from IBM, the settings for fairness are reversed.  Normally, fairness is turned               ON. However, if the OPTIONS diskette v1.0 is used,  "Fairness"
   needs to be set to OFF.  On all other versions of the OPTIONS diskette, simply set fairness to the normal setting of ON.

3. Verify the disk driver.  Verify with the third-party HBA
manufacturer that the latest Novell certified drivers are being
used and that it is compatible with the firmware revision           level on your HBA.

4. Verify the HBA firmware.  Verify with the third-party HBA manufacturer that the version of firmware on the HBA is the       correct version applicable to the version of the disk driver           that is being used.

5. Verify the firmware revision level on the drive.  Verify with       the third-party drive manufacturer that the proper or latest version of firmware exists the drive.

6. Check, and if necessary replace the cabling on the SCSI bus. Reseat all connectors, then reseat the HBA into the bus.  Look for pinched or broken leads on all connectors and cables.  Novell Labs in conjunction with HBA manufacturers has determined that some SCSI external cables are not designed properly for current applications.  Testing personnel were experiencing problems with drive deactivation. High speed testing equipment indicated that the cables were not designed  to specification.  Higher quality cables were then used and the drive deactivation ceased.
Experience has indicated that drive deactivation is more prevalent as the number of external devices increases and as the total length of the SCSI bus increases.  If the deactivation is limited to external devices, then be sure to use high quality cables, keep them as short as possible.  
Note: The SCSI spec for single-ended buses (which is what most all  applications use) calls out a maximum of 6 meters or about 20 feet.  However, many installations have reported problems with 3 or more devices and as short as 3 meters.  Internal cables (50 pin flat-ribbon cables) are much more stable than external cables, however, some cables have their connectors too close to each other.  No two connectors with devices attached should be closer than 1 foot.

7. Verify proper wattage of the power supply.  Many drive deactivation problems have been solved by upgrading or replacing a weak or flaky power supply.  Drive deactivation problems seem to be an ever increasing problem as drives grow larger in capacity yet shrink in physical size.  This smaller size allows more drives to be placed into the server box itself thus placing a larger demand on the server's power supply.

8. Examine other devices attached to the SCSI bus.  If a tape unit, CD-ROM, or other devices are attached to the SCSI bus, be sure to enable the "disconnect" feature on those devices.  This allows the HBA to issue a command or series of commands to that device and then disconnect and return to servicing disk requests.  If the "disconnect" feature is not enabled, the SCSI bus must wait for the device to return with a completion code and by so doing may not be able to service the drivers requests to "talk" to the drive.

9. Examine other bus mastering boards. If the server has bus mastering boards, they may need to be examined closely.  If a bus mastering board holds the bus too long, this may prevent the disk driver from establishing a communication link with the drive. Sometimes it may be an issue of getting a newer driver for the bus mastering device or taking drastic steps of replacing or upgrading the suspected bus mastering device.  Experience has shown that this can happen, but it is rare.

10.Replace the HBA. If the above steps have failed to solve the drive deactivation problem, then it is time to take drastic action.  Replace the HBA with preferably the same make and model number.  If a different brand or model number is chosen, a reformatting and repartitioning of the drive may be necessary.

11.If still unsuccessful in solving the drive deactivation problem, it is time to seriously look at replacing the drive.  Experience has shown that replacing the drive with the same make and model number may solve the problem; however, in some situations a switch to a different      manufacturer may be necessary.  Drive deactivation can sometimes be very finicky and may require drastic steps to stabilize the server and eliminate the drive deactivations.

Document Title:      Drive Deactivation Troubleshooting Tips
Document ID:      500089
Document Revision:      2
Creation Date:      18AUG1994
Modified Date:      20FEB1998
Novell Product Class:      NetWare

Novell Product and Version:      NetWare 3.11
      NetWare 3.12
      NetWare 3.2
      NetWare 4.01
      NetWare 4.02
      NetWare 4.1
      NetWare 4.11
      NetWare 4.2
      NetWare 5
      NetWare for Small Business 4.11
      NetWare for Small Business 4.2
      intraNetWare 4.11
      intraNetWare for Small Business 4.11

Free Tool: SSL Checker

Scans your site and returns information about your SSL implementation and certificate. Helpful for debugging and validating your SSL configuration.

One of a set of tools we are providing to everyone as a way of saying thank you for being a part of the community.

Just a comment:

Didn't look up your drive number but most of the IBM drives without termination are LVD drives and if your controller model number is correct, it doesn't support LVD. That means Single Ended Mode only!!

I've had problems like this. Most of the problems came from when setting termination, bad SCSI cables, cable length (total SCSI II + SCSI III) and !!BAD!! terminators. This problem reminds me of a Quantum drive that I had. It did the same thing. Replaced the drive and it was gone. This drive went crazy after functioning over a year!!

What I'm tring to say is;

1) Take your server apart. Look at your cables, jumpers, BIOS settings,..., have a cup of coffee and then look again. Get a friend to look at it. Make sure that the configuration conforms to the SCSI specs.

2) Replace the SCSI terminator.

3) Try trading out the drives (if you have a spare) one for one.

SCSI problems are a nightmare!

PS: According to the SCSI specs the TERMPWR may be turned on, on every device. Most Manufactures recommend that only one device supply. We have gone against their recomendations and have turned on the power when available for over 2 years now and haven't had !any! problems yet and this on over 100 servers! Don't think TERMPWR will solve your problem but it may help if you are pushing the length restrictions.
Chris RipollMaster of NoneCommented:
Just a stab in the dark !! What make/model is the computer (server) ?
I have exactly the same problem!

200 Bonus points for the person who comes up with the solution
We are still working on it so maybe i'll take them myself... ;)

Just came from the customer and tried some things, like added an external termintor to the bus.

BTW We had the "force SE" jumper set, but it still dismounted
the volumes.

The machine in question has a Chaintech BTS mainboard
(same adaptec controller onboard)


Will get back with more info when the server is still running in the morning.
Chris RipollMaster of NoneCommented:
I know this may sound very "untechnical" but I always try to eliminate the not-so-obvious. Do you have any power save options enabled in the bios ? Older SCSI drives didn't support power save, new ones do ??
Here we go:

To answer chrisrip's query first, all power save options have been turned off! But thanx for mentioning that, it CAN be the problem.

The machine we had the problems with now runs fine!

It seems there are 2 models drive with the same specs.
One auto terminated and one with NO termination.

We have the second one!

We put a passive, external terminator on the scsi bus which seems to have fixed the problem.

Also i removed jumper 6 "Force SE mode"

This doesn't seem to have much effect as the drive is supposed to auto detect the mode.

This is ONLY used if there are other SE drives on the same bus and you will want to force this drive to use that mode too.

Hope this will help solve your pronblem too davejhb

Question has a verified solution.

Are you are experiencing a similar issue? Get a personalized answer when you ask a related question.

Have a better answer? Share it in a comment.

Join & Write a Comment

Featured Post

Get expert help—faster!

Need expert help—fast? Use the Help Bell for personalized assistance getting answers to your important questions.

Tackle projects and never again get stuck behind a technical roadblock.
Join Now