Netservers
1748216 Members
3391 Online
108759 Solutions
New Discussion юеВ

Re: Proper way to replace hot swap drive on an LH3

 
Rob_171
Frequent Advisor

Proper way to replace hot swap drive on an LH3

Once before when I put in a new drive cage and set up a RAID 5 array I ran into a problem. One of the drives failed, and I replaced it, but I didn't have to be too careful since no data was on the drive. I messed around and finally got it working again.

What I want to know is, since I have all of my drives now working and storing data, is there a procedure documented on how to replace a failed drive? I get shivers at the thought of losing data, even though I have a backup in place (you all can see my view).

I have NetRaid Assistant installed, it's an older version that came with the Netserver back in 1998. I have no documentation on the software, so I was hoping to get a procedure written if ever needed in the future.

I have heard all I have to do is replace the current drive and it will automatically format and rebuild itself, but if that's the case, why does NetRaid Assistant give me the option to rebuild?

Any input is greatly appreciated. Thank you.
2 REPLIES 2
Rahul Sharma
Frequent Advisor

Re: Proper way to replace hot swap drive on an LH3

Normally you would log on to the box open compaq array utility(Now HP array utility). You should see the bad drive. From here you can remove or defunt the drive and then take out the drive.once you put in the new drive it will automatically start rebuilding. One thing i have learnt that once you have taken out a bad drive do not try to push it back again the server.I am not sure about your finding a documentation available for this.

kris rombauts
Honored Contributor

Re: Proper way to replace hot swap drive on an LH3

Hi Rob,

the previous post is mixing up with Proliant and SmartArray's as you could see and not Netservers/Netraid.


Now to your questions :

- a failed drive can be replaced online without any problem and a new disk can be inserted to replace it and a rebuild will take place automatically and when finished brings your array back in to redundant state where it can tolerate a new disk failure. No other user intervention is needed and server uptime or data is at all time available.

The way you get notified about a disk failure is (assuming you run Windows, if not pls specify) is by entries in the System Eventlog from different sources, depending on the fact if you installed the Netraid monitor software and eventual Netserver management agents (toptools agents, now called server agents).

A red led will also be flashing the disk when it is seen as failed by the Netraid controller.


You need to install a newer Netraid Assistant version, pls download it here :

http://h20000.www2.hp.com/bizsupport/TechSupport/SoftwareDescription.jsp?locale=en_USтМй=English&pnameOID=15389&prodSeriesId=50448&prodTypeId=15351&basePartNum=ns1868_1&locBasepartNum=ns1868_1en&os=Microsoft+Windows+2000&tech=Utility

and the monitor service :

http://h20000.www2.hp.com/bizsupport/TechSupport/SoftwareDescription.jsp?locale=en_USтМй=English&pnameOID=19774&prodSeriesId=74580&prodTypeId=15351&basePartNum=ns1961&locBasepartNum=ns1961en&os=Microsoft+Windows+2000+Server&tech=Utility++Storage+Controller

What you should do after installing this s/w is to make sure the weekly consistency check (a kind of health check ) is active on your server. This will do a complete check of the stripes in your raid5 every Sunday (default value). Pls check the readme files for details.


To answer the last question, their is always a manual way to start a rebuild for cases where users want to manually control this behavior and not start a rebuild of the array automatically (impacts the performance) during normal production hours.

HTH

Kris