Cable Select vs Master/Slave

darraghrogan

Hi,

Has anyone any experience of faulty HDs? Specifically, does whether a drive is CS or Master/Slave on an IDE chain decide whether the other drive will fail if the initial drive fails?

It's odd, but in a server in work, the HDs (two 120 Gb Maxtors) one a RMA model, the other about two years, are on the same IDE chain but the two year old one is making clunky noises intermittantly, and when it dies, so too does the newer one. Could this have anything to do with them both being CS?

Thanks,

Darragh

Capt'n Midnight

wouldn't think so.

on some old drive < 1GB some BIOS's would not detect the drive if it had been set as "master with slave present" and the slave was missing.

Just to be clear , is it the BIOS failing to see both drives when the older one dies, which would be strange unless the board on the drive was sending junk down the IDE cable. If on the other hand the OS fails to see both drives , that I'd expect as most OS's would get upset by a drive disappearing off a non-removeable interface.

time to get a new drive methinks, you did say it was a server, the data on it has just gotta be worth more than a new 120GB drive eg: it would cost the company less than €100 to have a new drive couriered on monday morning eg: http://www.jaguarcompsys.ie/components/index.htm#hd - €70 + VAT


if you wanted to waste time diaging ( on the off chance you could sort out a problem on a failing drive :rolleyes: ) you could change them to master / slave or just put the old drive on the other IDE cable and see if the CD or whatever keels over - but life's too short

galwaydude18

I Would Almost Be 110% Absolutly Sure It Is Not Do With Them Bing Cs! Why Dont You Set Them To Master And Slave Anyhow?

[EDIT] DAMN BEATIN TO IT[/EDIT]

darraghrogan

I'm RMAing the drive on Monday, and buying a new one at the same time to put it into the machine until RMA drive gets back.

THe reason I ask the question is from a server design standpoint. If one HD fails, the other one fails as well? Not too good considering that it could be a few days before I look at that server again. I'll try putting the HDs on different IDE chains..but would I risk cocking up the OS HD then?

Darragh

duridian

I think it may be more likely down to the fact that a lot of Maxtor hard drives manufactured up to about 2 years ago give problems, I'm not trying to Maxtor-bash here but I have found that a lot of the older ones give trouble. About 2 years ago Maxtor and Quantum joined forces to become one company and their more recent offerings seem to be better, fingers crossed that this will continue.

causal

With my puny understanding of IDE chains I thought:

If you have the two drives set as Master and Slave - and then the master fails - then the slave would no longer be detected, because Slave with no Master.
Similarly,
If you have the two drives set as Cable Select - and then the master fails - then the slave would no longer be detected, because Slave with no Master.

This would explain why when the master goes you lose the slave too.
Or am I just rounding off the square peg to make it fit the round hole!

hth,
causAl
(I've got jumpers on my pins)

tonky

Now that's creative !
It also probably contains the answer / (nods to Captn' Midnight) / possibly when the drive fails it pulls the M/S,CS,other pins to some default level, causing the second drive to disappear off the IDE bus.
Here's a couple of bits from my 'notes', authors unknown:

> The idea AFAIK is that unlike SCSI, IDE has the master drive control the
> slave drive on the same controller.
> 1) This causes problems where the drives are a different
> model/make/manufacturer as the master drive will downgrade the settings of > drives (master & slave) to the lowest common denominator; and therefore,
> speed in this case is affected.
> 2) Having the master drive control the slave drive also causes problems if
> the master fails as then both drives fail or if a drive takes down the chain
> both drives fail.

Also //

It is very important, that you only use one IDE disk per IDE bus. Not only would two disks ruin the performance, but the failure of a disk often guarantees the failure of the bus, and therefore the failure of all disks on that bus. In a fault-tolerant RAID setup (RAID levels 1,4,5), the failure of one disk can be handled, but the failure of two disks (the two disks on the bus that fails due to the failure of the one disk) will render the array unusable. Also, when the master drive on a bus fails, the slave or the IDE controller may get awfully confused. One bus, one drive, that's the rule.

darraghrogan

tonky wrote:

Now that's creative !
It also probably contains the answer / (nods to Captn' Midnight) / possibly when the drive fails it pulls the M/S,CS,other pins to some default level, causing the second drive to disappear off the IDE bus.
Here's a couple of bits from my 'notes', authors unknown:

> The idea AFAIK is that unlike SCSI, IDE has the master drive control the
> slave drive on the same controller.
> 1) This causes problems where the drives are a different
> model/make/manufacturer as the master drive will downgrade the settings of > drives (master & slave) to the lowest common denominator; and therefore,
> speed in this case is affected.
> 2) Having the master drive control the slave drive also causes problems if
> the master fails as then both drives fail or if a drive takes down the chain
> both drives fail.

Also //

It is very important, that you only use one IDE disk per IDE bus. Not only would two disks ruin the performance, but the failure of a disk often guarantees the failure of the bus, and therefore the failure of all disks on that bus. In a fault-tolerant RAID setup (RAID levels 1,4,5), the failure of one disk can be handled, but the failure of two disks (the two disks on the bus that fails due to the failure of the one disk) will render the array unusable. Also, when the master drive on a bus fails, the slave or the IDE controller may get awfully confused. One bus, one drive, that's the rule.

I don't like this but I suppose I'll have to live with it! THe mobo also has a RAID controller for IDE devices which I'm not using at the moment - mainly because I can't use a SMART monitoring app to keep an eye on RAID disks. If I put one Maxtor on it's own IDE chain, one on the RAID chain, would the devices be resiliant enough to operate independantly?

Does SATA have this problem?

Darragh

Mutant_Fruit

Master/Slave has no effect on the drives performance at all. As far as i know, the only difference it makes is for the OS. Windows will assign drive letters to the master drive first, and the master drive gets "chosen" first, if you know what i mean.

Currently i have RMA'd my "master" HD and now i only have my "slave" HD, which has windows on it. THe drive is still being recognised as the slave even though there is no master drive, and my DVD Burner has always been recognised as slave, as its set to slave. And everything works perfectly.

IF they both stopped being recognised, its possible the dead drive is fecking up data on the IDE cable. But it won't kill your good drive out of spite And the good drive won't kill itself out of sympathy

1) This causes problems where the drives are a different model/make/manufacturer as the master drive will downgrade the settings of drives (master & slave) to the lowest common denominator; and therefore, speed in this case is affected.

Thats only referring to the DMA setting. If you have a HD that supports DMA mode 33, 66 and 100, and a HD that supports DMA mode 33 and 66, when they are both connected to the same cable, the cable runs at the fastest speed thats supported by both drives. If it didn't, the slower drive would corrupt.

duridian

What Mutant Fruit said is right, having a faulty drive and a good drive together as master and slave or vice-versa doesn't mean the second drive will become faulty as well. Errors caused by the faulty drive may also cause recognition of a good drive on the same IDE cable to fail, but it doesn't mean anything is wrong with that drive, and, on it's own the good drive will continue to work fine.
Also in a cable select setup the priority of the drives as master/slave is determined by which of the connectors on the IDE cable you connect the drive. Hence the term 'cable select'. Normally the one furthest from the motherboard end of the cable becomes master.

Gone West

I didnt think m/s or c/s had any effect on hd failure?