I'm not familiar with the HS22, but we have a Dell Blade Centre, and I'm guessing that the reasons will be very similar. The Dell can actually take up to 5 minutes to start booting the OS.
To start with, in a blade centre, everything is high quality, and high quality parts have always taken a long time to POST themselves. Usually the memory check when the server boots up is not set to 'quick', and if there's a lot of RAM it can take a long time to check itself.
The on-board SCSI/SATA/SAS/RAID always takes a long time to boot. It needs to POST itself, and then do a scan to find all the attached disks. Once it finds all the attached disks, it needs to read the metadata and create the arrays. This is always slow. Maybe 60-90 seconds just for this alone.
Additional overheads with blades may be the blade registering itself with any switches or FC-AL cards that are plugged in.
None of these things you usually have any control over, and if you do, the fact that they take a long time is usually a good thing. Doing a soft-reboot is usually much, much faster than a cold-boot.
all new servers take a long time to boot. the reason is - SAS drives.
while the old scsi drives were enumerated according to their location on the bus, the SAS drives are noted by the controller by their unique WWN, which means they all must be scanned, and tested for raid config presence. The advantage is - the drives can move around the backplane with no issues, and raid configurations can be migrated with more ease.
plus the serial vs parallel speed difference of course
I'm not familiar with the HS22, but we have a Dell Blade Centre, and I'm guessing that the reasons will be very similar. The Dell can actually take up to 5 minutes to start booting the OS.
To start with, in a blade centre, everything is high quality, and high quality parts have always taken a long time to POST themselves. Usually the memory check when the server boots up is not set to 'quick', and if there's a lot of RAM it can take a long time to check itself.
The on-board SCSI/SATA/SAS/RAID always takes a long time to boot. It needs to POST itself, and then do a scan to find all the attached disks. Once it finds all the attached disks, it needs to read the metadata and create the arrays. This is always slow. Maybe 60-90 seconds just for this alone.
Additional overheads with blades may be the blade registering itself with any switches or FC-AL cards that are plugged in.
None of these things you usually have any control over, and if you do, the fact that they take a long time is usually a good thing. Doing a soft-reboot is usually much, much faster than a cold-boot.
all new servers take a long time to boot. the reason is - SAS drives. while the old scsi drives were enumerated according to their location on the bus, the SAS drives are noted by the controller by their unique WWN, which means they all must be scanned, and tested for raid config presence. The advantage is - the drives can move around the backplane with no issues, and raid configurations can be migrated with more ease. plus the serial vs parallel speed difference of course