+ -2 on RAID 5
On Thu, May 1, 2014 at 8:02 PM, Micheal Espinola Jr < [email protected]> wrote: > Absolutely. But you dont want to make this discovery mid-procedure, when > you've already started swapping drives. Older drives will be closer to > their MTF. The more you physically interact with said drive, the more > causality your are exposing yourself to induce a failure. > > As someone that has performed these tasks many times, I can say from > experience that drives can fail simply by moving them. > > So my recommendation is based on performing a task that would give you the > least exposure to possible critical failures - regardless of likelihood. > > -- > Espi > > > > On Wed, Apr 30, 2014 at 3:15 AM, Ken Schaefer <[email protected]> wrote: > >> If your backup/restore process doesn’t work, this is probably one of >> the better times to find out (since the old disks are available and have >> current data). Far better to find out now that it doesn’t work, rather than >> when you really need it. >> >> >> >> Cheers >> >> Ken >> >> >> >> *From:* [email protected] [mailto: >> [email protected]] *On Behalf Of *Micheal Espinola Jr >> *Sent:* Tuesday, 29 April 2014 10:48 AM >> >> *To:* ntsysadm >> *Subject:* Re: [NTSysADM] RE: server migration w/new HDs >> >> >> >> Backup/restore situations are often ones that bite people in the ass when >> they least expect it to. In this situation, I would not opt to put that to >> the test unless it was absolutely a neccessity. >> >> >> -- >> Espi >> >> >> >> >> >> On Mon, Apr 28, 2014 at 3:01 PM, Kent McKinney <[email protected]> >> wrote: >> >> Even if you swap one drive at a time and get the virtual disk size >> increased, the OS partition would also need to be resized. Its time >> consuming and will greatly affect server performance. Safest way is to >> install the server backup feature, do a system state backup, pull the old >> drives out (with existing data intact that you can always roll back to) >> create the new array and restore. >> >> --- Original Message --- >> >> From: "Maglinger, Paul" <[email protected]> >> Sent: April 28, 2014 5:34 PM >> To: "'[email protected]'" <[email protected]> >> Subject: [NTSysADM] RE: server migration w/new HDs >> >> This is a multi-part message in MIME format. >> >> >> --_000_36DA94300D69184A91500B695547A148481711CCCOMSTAR1scvlcom_ >> Content-Type: text/plain; charset="us-ascii" >> Content-Transfer-Encoding: quoted-printable >> >> >> I just talked with someone about a similar situation. >> >> HP-UX server running 2 72G mirrored drives and had one fail. The vendor >> su= >> bbed a 146G drive. In this case it would rebuild to only 72G of the 146 >> dr= >> ive, but the vendor said that if the other 72 drive failed and was >> replaced= >> >> with the 146 then I could extend it to the full 146 at that point >> . >> >> I have not tested this, and am not sure if it would work, but wondering >> if = >> you could replace the drives one at a time allowing for the rebuild. I >> did= >> something similar with a EVA SAN and was successful. It took a long >> time,= >> but I didn't have to take the system down. I'm not sure I'd risk it on >> a = >> >> production system. >> Can you do a system level backup/restore? >> >> From: [email protected] [mailto: >> [email protected]= >> >> ] On Behalf Of Jesse Rink >> Sent: Monday, April 28, 2014 4:07 PM >> To: [email protected] >> Subject: [NTSysADM] server migration w/new HDs >> >> Have a Dell T310 server with (4) 1TB sas drives in a RAID5 array, >> running W= >> indows 2008R2. Need to replace the drives with (4) 2TB sas drives in a >> RAI= >> D5 array. Server only holds 4 physical HD bays so I can't just add an >> ex= >> tra array. Curious if anyone has any tricks to speeding up this >> migration = >> other than rebuilding the OS from scratch on the (4) new 2Tb sas drives >> and= >> reinstalling all the app >> >> s on the server, etc. >> >> >> >> JR >> >> >> >> >> >> > >
