BladeSystem - General
1748054 Members
4891 Online
108758 Solutions
New Discussion юеВ

Re: 4 minutes boot time - is this normal?

 

4 minutes boot time - is this normal?

I'm quite new to the world of HP blades, and can't help being a bit surprised about the time it takes to power on and boot our BL460c G7 blades with 96GB RAM - fresh from the factory. I thought our IBM and Sun servers were slow starters, but oh man:

(min:sec)
00:00 Power-on
01:38 First video sync on the iLO (ProLiant logo)
02:04 Controller Status: Init done
03:15 Configuring ServerEngines Controller...
04:10 Starting boot of OS

Any settings available to speed this up, or is this just how it is supposed to be? The blade doesn't seem to be running any extended diagnostics that would warrant this.
4 REPLIES 4
JKytsi
Honored Contributor

Re: 4 minutes boot time - is this normal?

New servers do more self diagnostics than older machines. Also checking the memory will take time.

Is it always same time ? No POST messages ?
Remember to give Kudos to answers! (click the KUDOS star)

You can find me from Twitter @JKytsi
kent as
Occasional Advisor

Re: 4 minutes boot time - is this normal?

Hi,
We do experience same thing here. Ridicules long boot time on our Bl465 G7 servers. Three minutes and thirty seconds to get the operating system start loading feels like forever when you├в re forced to reboot a server in production.
Gerardo Arceri
Trusted Contributor

Re: 4 minutes boot time - is this normal?

I was wondering about the same
we have a mix of BL680c G5's / BL460 G6's and now BL460 G7, so far i always complained about how slow the 680's boot process was, and i had seen boot times improve with BL460 G6s and get even worse with G7s, what's the deal here ?
Why run more diags in one generation than the other ?
JonathanT
Frequent Advisor

Re: 4 minutes boot time - is this normal?

Your newer servers are going to typically have more RAM to check, etc. You want your servers to do diagnostics to catch any kind of hardware problems up front. Consider virtualizing a lot of your workload so that if you need to reboot a physical server you can shift the running virtual machines to other hardware and perform the reboot without clients noticing any downtime.