Jump to content

Time to Boot?


RandomBox

Time to Boot?  

27 members have voted

You do not have permission to vote in this poll, or see the poll results. Please sign in or register to vote in this poll.

Recommended Posts

I have an Adaptec U160 SCSI controller (w/6HDD @ 10k+RPM) that takes longer to boot/initialize than the normal BIOS.But nooooooooooooooooo, you did not see me use that as an excuse for my pathetic 143 second result! icon8.gif--------------------------------------------Let the above "RPM" statement be a lesson to you burninbush.Fast rotational-speed HDDs are great to read/write to/from but they really are not going to be the panacea that you expect for your boot times: Technically during boot, a faster RPM HDD is going to take longer time to spool up and/or stabilize.But to listen to the starting whine of a 15k RPM HDD (Fujitsu MAS3735NP w/3ms access time) is such a delight. thumbsup.gif+++++++++++++Thanks, but no lesson is needed here. I've been working with various sorts of disk drives professionally for the last 20 years or so. The problem with this particular sub-topic is that there are two specialized types of disks being tossed into the mix with disk drives in general. 10k and 15k scsi drives are special cases, and they do take longer to rev-up, while 4200rpm and 5400rpm laptop drives are also special cases, they are built to use as little power as possible so they also take a little longer to spin up. But there are a few 7200rpm laptop drives that start as fast as the typical desktop ide disk -- D*** the power -- and that is what I propose to use as a replacement. Even if they were proportionately slower [like the scsi disks] the difference in spin up would only be a second or two. 7200rpm IDE desktop disks don't start instantly either. The main reason a 7200 drive would boot faster is that it would load the OS code to memory a lot faster -- more sectors passing under the heads every second. This is not insignificant -- the 7200 is 33% faster even at the same size. Actually, all I opined was that it would cut a couple seconds from the HP bootup time. The "143 seconds" time noted for 15k scsi up the thread is not the disk's spin up time -- the bulk of the =extra= scsi delay is the time it takes the ultra-scsi controller to scan for 15 possible additional devices -- and timeout on those not attached. I'm very familiar with this, running from an Adaptec u2w card and a couple 10k drives at this moment; that card adds about 45 seconds to every bootup, either win2k or linux, half of that in bios POST and the other half when the OS is coming up and does the same devices scan again. Again, the disk spin up time is only a small fraction of the total. My guess is that the 143 seconds time reported is the total bootup = OS + scsi disk times. Probably 2/3rds of that is in the bios and the OS, after the disks have been recognized.For anyone doubting the innate slowness of laptop drives, I suggest that you get a 44-to-40 adapter and connect one as a slave in your desktop -- see the difference for yourself. It's very perceptible. Windows users can get a detailed disk speed test with the hdtach-261 util; linux users have hdparm, but that seems to be incredibly optimistic, only tests at the beginning of the disk, and the DMA speed reported is surely from memory buffer to memory -- for example: root@xp2400-knopp:~# hdparm -tT /dev/hda/dev/hda: Timing cached reads: 1020 MB in 2.00 seconds = 509.59 MB/sec Timing buffered disk reads: 176 MB in 3.02 seconds = 58.36 MB/secroot@xp2400-knopp:~#The /dev/hda above is a 250gb ata100 Maxtor ide.

Link to comment
Share on other sites

I just ran that post#5 referenced BootVis utility on my system.BootTrace function within BootVis considers my systems Boot Time ("Boot Done") to be 73.80Seconds.It further shows that the last process created ("Process Creates") was at 108.39Seconds.But the BootVis.exe process didnot start until 138.04Seconds.138.04Seconds is about when all of the clatter from the drives (diskkeeper, avg, etc) stop.But in all honesty, it must be said that I do run a loaded system at bootup. 14 items in TaskTray alone! This is NOT an excuse and I really don't consider my system to be slow at all. It is just 'burdeneded' with many toolz that I consider worthy to run at all times. Heck my FireFox alone has 45 extensions alone. :devil:BootVis BootOptimize function next! B)

Link to comment
Share on other sites

The "143 seconds" time noted for 15k scsi up the thread is not the disk's spin up time -- the bulk of the =extra= scsi delay is the time it takes the ultra-scsi controller to scan for 15 possible additional devices -- and timeout on those not attached. I'm very familiar with this, running from an Adaptec u2w card and a couple 10k drives at this moment; that card adds about 45 seconds to every bootup, either win2k or linux, half of that in bios POST and the other half when the OS is coming up and does the same devices scan again. Again, the disk spin up time is only a small fraction of the total.
I had scsi (old style scsi with only 8 scsi id's) in a couple of old Amiga's.Don't recall just which file I used edited to force the controller to only look for a specific subset of scsi id's (of devices actually attached to the system), but it made a HUGE difference in the boot time.No clue if it's even possible to do something similar under Windows/Vista.
Link to comment
Share on other sites

... Don't recall just which file I used edited to force the controller to only look for a specific subset of scsi id's (of devices actually attached to the system), but it made a HUGE difference in the boot time. No clue if it's even possible to do something similar under Windows/Vista.
There are a bunch of such commands in newer versions of SCSI BIOS. I think you may be referring to either one of these:*Send Start Unit Command (per ID#)*Include in BIOS Scan (per ID#)The newer SCSI U160/U320 Cards can control upto 15 devices per channel and some controllers are Dual Channel.
Link to comment
Share on other sites

To address just the items in the last two posts, -- re the scsi IF searching all possible units -- The setting to exclude devices from the scan is in the Adaptec bios add-on for those cards I'm familiar with -- do a ctrl-A early in the bios POST, when the Adaptec logo shows, and that brings up the card's setup screen. Sending scsi commands to the card might be useful while the OS is booting up, after the kernel has loaded, but bios doesn't have that option. The wide [16-bit] scsi cards have 4-bit addressing, and can address 16 devices on one cable. The card itself uses one address, so you can have 15 disks or tapes or cdroms or scanners or whatever else on the cable in a long daisy-chain. Even if you disable the scan for all devices you aren't using, it still does some sort of a lookup on them. I shouldn't have used the word "timeout" -- that's probably not exactly what's going on in this detailed situation. Dunno what you really should call it, but even with all devices excluded, the card takes a good long time to pass on to the rest of the POST sequence. And, what seems really dumb to me, when the OS boots up you get the same delay again, a double whammy. SCSI is a PITA, generally, IMO [LOL] on home computers. The same basic drive will cost more with a scsi interface, and the cabling is horribly expensive, and you have to buy that pci card and a terminator, not cheap either. I wouldn't be using scsi except that I had a source for more free disks than I could ever use from my last job -- we upgraded our users [sparc stations] with new disks just about every year, and they'd just wipe and toss the removed drives. I was in position to catch the toss, as many as I'd want. Only hitch was that they were all SCA [another acronym] -- so I had to buy adapters to connect them to ordinary cables. Did I mention that everything about scsi is expensive? There's no way this would be economical for a home desktop user with one or two disks. If you're running a database where lots of users look up small amounts of data then a fast scsi disk is the hot setup. Continuous high-speed seeking is what they are best at. For continuous data xfer rates, the same equation holds for both the scsi and the ide disks -- sectors passing under the active head in unit time.

Link to comment
Share on other sites

Guest LilBambi

On the new Mac Mini G4 Power PC (1.42Ghz) reboot takes about 26-30 seconds ... not totally sure because I sat there in shock at how fast it booted for 2-4 seconds! LOL!

Link to comment
Share on other sites

:hysterical: lilbambi, i sit in shock when some of my equipment even bothers to boot! :hysterical:
Yeah, especially knowing that you had turned it OFF the night before and it was in perfect working condition... grrrrr.gifOne of the reasons I kinda get a kick out of WinOSs: They have a weird character ...all their own
Link to comment
Share on other sites

  • 3 weeks later...

WinXp SP2 - Intel 2.6ghz - 500MB RAM - AVG - Mailwasher - Soundmax - DHCP - Execute Windows Explorer Etc. = 55 seconds.Shutdown 12 Seconds.

Edited by Ozidave
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...