SCSI transport error

[NavinB]

Hi Everyone,

I am getting this error message on my sun server Ultra 10:-

Feb 23 09:44:51 ospf unix: WARNING: /sbus@1f,0/SUNW,fas@e,8800000 (fas0):
Feb 23 09:44:51 ospf data transfer overrun: current fas state:
Feb 23 09:44:51 ospf unix: fas: Latched stat=0x90<IPND,XZERO> intr=0x10<BUS>
Feb 23 09:44:51 ospf unix: fas: last msgout: SIMPLE QUEUE TAG, last msgin: CO
AND COMPLETE
Feb 23 09:44:51 ospf unix: fas: DMA csr=0xb2bf8810<TWOCYC,BRST1,BST0,DSBLEDRN
IDE,INTEN>
Feb 23 09:44:51 ospf unix: fas: addr=0 dmacnt=0 test=0 last=f0518000 last_cnt
000
Feb 23 09:44:51 ospf unix: fas: fas state:
Feb 23 09:44:51 ospf unix: fas: count(32)=0 cmd=1 stat=13 stat2=80 in
=0
Feb 23 09:44:51 ospf unix: fas: step=cc fifoflag=80 conf=97 test=1e c
f2=90 conf3=c3
Feb 23 09:44:51 ospf unix: fas: Cmd dump for Target 8 Lun 0:
Feb 23 09:44:51 ospf unix: fas: cdb=[ 0x2a 0x0 0x1 0x8d 0x19 0x80 0x0 0x0 0x
0x0 ]
Feb 23 09:44:51 ospf unix: fas: State=DATA Last State=DATA_DONE
Feb 23 09:44:51 ospf unix: fas: pkt_state=0xf<XFER,CMD,SEL,ARB> pkt_flags=0x4
0 pkt_statistics=0x2
Feb 23 09:44:51 ospf unix: WARNING: /sbus@1f,0/SUNW,fas@e,8800000 (fas0):
Feb 23 09:44:51 ospf Target 8 reducing sync. transfer rate
Feb 23 09:44:51 ospf unix: WARNING: /sbus@1f,0/SUNW,fas@e,8800000/sd@8,0 (sd7):
Feb 23 09:44:51 ospf SCSI transport failed: reason 'data_ovr': retrying co


Any idea,what it is and how to avoid this ....

Thanks in anticipation

 

[cndcadams]

Well this error Target 8 reducing sync. transfer rate

Things that can cause this.

1 you have drives on the same chain that are different RPM.
The faster drive with reduce transfer rate to the slowest drive.

2 the drive is going bad. Most likely if the setup has been up and functional for a while and you have never seen these errors.

3. Termination can cause very strange issues.

If this was the only error I would continue to monitor drive Target8 /sbus@1f,0/SUNW,fas@e,8800000/sd@8,0 (sd7):

Be careful though, if there are errors before these lines you must make sure to determine the drive that first gave the error. Alot of the time a drive will give an error such as this and it will cause the drives on the same bus to display errors as the bus is resetting.

Hope this helps

Chris Adams

 

[NavinB]

Thanks for u'r reply.....

My prob is the first one that u hv described.....
The SCSI bus has a external disk and a tape drive which I added recently.....Is there any command to reduce data transfer rate of tape drive......????

Hoping to get a reply soon....

 

[Chacalinc]

No, there is no command. To avoid this issue, put the tape drive in a different SCSI bus. It is NOT recommended to share the same scsi bus between disk drives and tapes.

Tha main problem about sharing is: when a tape is doing rewind or forwarding, SCSI bus is blocked until the activity finish, avoiding data tranfers to the disks.

Cheers.

 

[NavinB]

Well I have only one SCSI interface on my sun server Ultra 10...What should I do in this case ?????
One more thing,I had a tape drive earlier also connected to the same SCSI bus..I just replaced it with a new one...and it was working fine before.....

 

[mikeclark]

Was the tape drive the same type. If this was replaced with a different drive and the system was not reconfigured this could be your problem.

 

[NavinB]

The tape drive is different from earlier one but is supported by Solaris 5.7 as well as my server.....

 

[cndcadams]

Is this tape drive a sun supported tape unit, What is the part#, also have you checked your /kernel/drv/st.conf file to see if there is an entry for your tape drive. In older OS if a tape drive was not in the file you would receive similar errors. There are also patches for solaris 7 st stuff.

Thanks

Chris A

 

[NavinB]

The tape drive is HP external drive and supported by solaris but there is no entry for this in /kernel/drv/st.conf.Shall I make an entry for tape drive in this file..what should the entry be like ???? or do I have to apply patch ????

Thanx....

 

[cndcadams]

I would put an entry in, I have done this in 2.6 years ago and had no problems. A little playing around. Just make sure to make a copy of your st.conf before editing it.You could try the patch, I believe you were running 2.7 patch is 107460-13. This patch has some fixes for STK and DDS stuff. You should look at the model of your old tape drive and see if that was in st.conf. I have a feeling that your problems are caused by the lack of an entry for the tape drive. Can you get the model# of the tape unit, maybe it is in solaris 9 st.conf which I have at home, I could get you the entry needed.

Thanks

Chris A.

 

[NavinB]

thanx for the reply Chris A...I'll definitely try this putting an entry and see what happens....

 

[NavinB]

Hi Chris !!!

Can u please send me the entry for the st.conf file as u had said.....????

Thanx in advance

 

[daFranze]

@NavinB

I would patch first! Some guys wrote books about the "right" st.conf entries...

Regards
-- Franz
Sorry I'm not a native spaeker, I'm from Munich, Germany - &quot;Home of the Whopper&quot;, oh no, &quot;Home of the Oktoberfest&quot; ;-)
Solaris System Manager; I used to work for Sun Microsystems Support (EMEA) for 5 years

 

[cndcadams]

Hey;

Sorry took me so long to get back to you. I have been swamped. Email me at cndcadams@aol.com and I will attach the file.

Thanks

CA