![]() | |
![]() |
| | Thread Tools | Display Modes |
#1
| |||
| |||
|
#2
| |||
| |||
|
|
I'm considering the use of the built-in RAID 1 feature (mirroring) for data reliabilty under Mac OS X for a FileMaker database. Does anyone know exactly what happens when one of the drives fails? I assume that, since it's RAID, the system should keep running transparently. So, how do we know the drive has failed? Does the RAID software give a notification (email, applescript, etc.) that will be noticed if one is not physically at the computer? |
#3
| |||
| |||
|
|
In article <1fz9ogj.1gpy6h7avkuvN%md03NOSPAM (AT) xochiNOSPAM (DOT) com.invalid>, md03NOSPAM (AT) xochiNOSPAM (DOT) com.invalid (Mike D) wrote: I'm considering the use of the built-in RAID 1 feature (mirroring) for data reliabilty under Mac OS X for a FileMaker database. Does anyone know exactly what happens when one of the drives fails? I assume that, since it's RAID, the system should keep running transparently. So, how do we know the drive has failed? Does the RAID software give a notification (email, applescript, etc.) that will be noticed if one is not physically at the computer? I have limited experience with this, so take it as a data point and nothing more. we're running the built-in Raid feature (mirroring) on an XServe. Those of you familiar with the hard drive bays on the XServe know that they just need a little push and they pop out. Someone bumped a chair into one of the hard drives and pushed it out. The server continued to run fine on one drive from that point on (although in a "degraded" state) and it was some time before we noticed. When we pushed the hard drive back in, we had to go back to Disk Utility to resync the two drives to re-establish the RAID. If we had configured Server Monitor to notify us, it could have sent us email or paged us alerting us to the degraded array (we subsequently set it up to do that). However, Server Monitor is a OS X Server utility; not sure if something like it exists for plain ol' OS X. I only noticed when I happened to see that both little drive access lights weren't blinking in unison like they usually did. |
#4
| |||
| |||
|
|
It sounds like my concern is valid: that unless running OSX Server with Server Monitor, one could lose one drive an never know it untill the 2nd drive failed as well... |
#5
| |||
| |||
|
|
sam grey <sgrey (AT) invalid (DOT) com> wrote: In article <1fzbiu7.16dsvyh1u4on4qN%md03NOSPAM (AT) xochiNOSPAM (DOT) com.invalid>, md03NOSPAM (AT) xochiNOSPAM (DOT) com.invalid (Mike D) wrote: Thank you! It sounds like my concern is valid: that unless running OSX Server with Server Monitor, one could lose one drive an never know it untill the 2nd drive failed as well... I suppose; you could keep checking with Disk Utility, which tells you when and if an array is degraded. Interesting idea. I wonder if there is a command-line utility to do this? Any darwin/unix gurus out there? If there is, it should be trivial to write a cron job to send an email when one drive fails... |
#6
| |||
| |||
|
|
In article <1fzd4sm.8kei8511voq6nN%md03NOSPAM (AT) xochiNOSPAM (DOT) com.invalid>, Interesting idea. I wonder if there is a command-line utility to do this? Any darwin/unix gurus out there? If there is, it should be trivial to write a cron job to send an email when one drive fails... The command line version of Disk Utility is called "diskutil." You probably want either the "info" subcommand or "checkRAID" -- I bet the latter. |
![]() |
| Thread Tools | |
| Display Modes | |
| |