* RAID HDDs spin up sequence @ 2011-01-31 20:18 Piergiorgio Sartor 2011-01-31 21:09 ` Roberto Spadim 0 siblings, 1 reply; 22+ messages in thread From: Piergiorgio Sartor @ 2011-01-31 20:18 UTC (permalink / raw) To: linux-raid Hi all, assuming there is a NAS, with, for example, 10 HDDs in RAID-6. Assuming the HDDs are put in standby, in order to save energy, when the NAS is not used. How is the spin up sequence when the corresponding /dev/mdX device is accessed? Will the system spin up one HDD at time or all together? Obviously, one at time will limit the peak current, thus allowing a better dimensioned PSU, working almost always around the optimal efficiency point. Thanks a lot for any information on the topic, bye, -- piergiorgio ^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: RAID HDDs spin up sequence 2011-01-31 20:18 RAID HDDs spin up sequence Piergiorgio Sartor @ 2011-01-31 21:09 ` Roberto Spadim 2011-01-31 21:10 ` Roberto Spadim ` (2 more replies) 0 siblings, 3 replies; 22+ messages in thread From: Roberto Spadim @ 2011-01-31 21:09 UTC (permalink / raw) To: Piergiorgio Sartor; +Cc: linux-raid you psu must be dimensioned to work with everythink at full work load (it´s a real production NAS right?! not a test) your SAS/IDE/SATA controller and HDD manual should be checked how hdd wake up? one command (read/write) over sata/sas/ide channel wake it up? on linux raid we have a read algorithm and a write algorithm if a raid1 write occur all disks will wake up if a raid1 (raid0 or another) read occur only the disk will wake up but check you SATA/IDE/SATA controller, how it wake up your disk, and how you hdd wake up 2011/1/31 Piergiorgio Sartor <piergiorgio.sartor@nexgo.de>: > Hi all, > > assuming there is a NAS, with, for example, 10 HDDs > in RAID-6. Assuming the HDDs are put in standby, in > order to save energy, when the NAS is not used. > How is the spin up sequence when the corresponding > /dev/mdX device is accessed? > Will the system spin up one HDD at time or all together? > > Obviously, one at time will limit the peak current, > thus allowing a better dimensioned PSU, working almost > always around the optimal efficiency point. > > Thanks a lot for any information on the topic, > > bye, > > -- > > piergiorgio > -- > To unsubscribe from this list: send the line "unsubscribe linux-raid" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html > -- Roberto Spadim Spadim Technology / SPAEmpresarial -- To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html ^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: RAID HDDs spin up sequence 2011-01-31 21:09 ` Roberto Spadim @ 2011-01-31 21:10 ` Roberto Spadim 2011-01-31 21:11 ` Mathias Burén 2011-01-31 22:23 ` Piergiorgio Sartor 2 siblings, 0 replies; 22+ messages in thread From: Roberto Spadim @ 2011-01-31 21:10 UTC (permalink / raw) To: Piergiorgio Sartor; +Cc: linux-raid normaly... raid1 send many writes and one read raid0 send one write and one read 2011/1/31 Roberto Spadim <roberto@spadim.com.br>: > you psu must be dimensioned to work with everythink at full work load > (it´s a real production NAS right?! not a test) > your SAS/IDE/SATA controller and HDD manual should be checked > how hdd wake up? one command (read/write) over sata/sas/ide channel wake it up? > on linux raid we have a read algorithm and a write algorithm > if a raid1 write occur all disks will wake up > if a raid1 (raid0 or another) read occur only the disk will wake up > > but check you SATA/IDE/SATA controller, how it wake up your disk, and > how you hdd wake up > > 2011/1/31 Piergiorgio Sartor <piergiorgio.sartor@nexgo.de>: >> Hi all, >> >> assuming there is a NAS, with, for example, 10 HDDs >> in RAID-6. Assuming the HDDs are put in standby, in >> order to save energy, when the NAS is not used. >> How is the spin up sequence when the corresponding >> /dev/mdX device is accessed? >> Will the system spin up one HDD at time or all together? >> >> Obviously, one at time will limit the peak current, >> thus allowing a better dimensioned PSU, working almost >> always around the optimal efficiency point. >> >> Thanks a lot for any information on the topic, >> >> bye, >> >> -- >> >> piergiorgio >> -- >> To unsubscribe from this list: send the line "unsubscribe linux-raid" in >> the body of a message to majordomo@vger.kernel.org >> More majordomo info at http://vger.kernel.org/majordomo-info.html >> > > > > -- > Roberto Spadim > Spadim Technology / SPAEmpresarial > -- Roberto Spadim Spadim Technology / SPAEmpresarial -- To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html ^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: RAID HDDs spin up sequence 2011-01-31 21:09 ` Roberto Spadim 2011-01-31 21:10 ` Roberto Spadim @ 2011-01-31 21:11 ` Mathias Burén 2011-01-31 21:25 ` Roberto Spadim 2011-01-31 22:23 ` Piergiorgio Sartor 2 siblings, 1 reply; 22+ messages in thread From: Mathias Burén @ 2011-01-31 21:11 UTC (permalink / raw) To: Roberto Spadim; +Cc: Piergiorgio Sartor, linux-raid On 31 January 2011 21:09, Roberto Spadim <roberto@spadim.com.br> wrote: > you psu must be dimensioned to work with everythink at full work load > (it´s a real production NAS right?! not a test) > your SAS/IDE/SATA controller and HDD manual should be checked > how hdd wake up? one command (read/write) over sata/sas/ide channel wake it up? > on linux raid we have a read algorithm and a write algorithm > if a raid1 write occur all disks will wake up > if a raid1 (raid0 or another) read occur only the disk will wake up > > but check you SATA/IDE/SATA controller, how it wake up your disk, and > how you hdd wake up > > 2011/1/31 Piergiorgio Sartor <piergiorgio.sartor@nexgo.de>: >> Hi all, >> >> assuming there is a NAS, with, for example, 10 HDDs >> in RAID-6. Assuming the HDDs are put in standby, in >> order to save energy, when the NAS is not used. >> How is the spin up sequence when the corresponding >> /dev/mdX device is accessed? >> Will the system spin up one HDD at time or all together? >> >> Obviously, one at time will limit the peak current, >> thus allowing a better dimensioned PSU, working almost >> always around the optimal efficiency point. >> >> Thanks a lot for any information on the topic, >> >> bye, >> >> -- >> >> piergiorgio >> -- >> To unsubscribe from this list: send the line "unsubscribe linux-raid" in >> the body of a message to majordomo@vger.kernel.org >> More majordomo info at http://vger.kernel.org/majordomo-info.html >> > > > > -- > Roberto Spadim > Spadim Technology / SPAEmpresarial > -- > To unsubscribe from this list: send the line "unsubscribe linux-raid" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html > I would guess, that on a RAID0 setup, any read to that md device would wake every disk up in that setup. No? // Mathias -- To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html ^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: RAID HDDs spin up sequence 2011-01-31 21:11 ` Mathias Burén @ 2011-01-31 21:25 ` Roberto Spadim 2011-01-31 21:29 ` Mathias Burén 0 siblings, 1 reply; 22+ messages in thread From: Roberto Spadim @ 2011-01-31 21:25 UTC (permalink / raw) To: Mathias Burén; +Cc: Piergiorgio Sartor, linux-raid just if you need information on all disks define your read. are you reading 1 byte or many bytes (on all hds?) the number of bytes read/write is the point if you need another disk to read/write you need to wake up your another hd check this implementations on mdadm: linear, raid0, 0, stripe maybe one is better for low power than another, but maybe one is faster than another 2011/1/31 Mathias Burén <mathias.buren@gmail.com>: > On 31 January 2011 21:09, Roberto Spadim <roberto@spadim.com.br> wrote: >> you psu must be dimensioned to work with everythink at full work load >> (it´s a real production NAS right?! not a test) >> your SAS/IDE/SATA controller and HDD manual should be checked >> how hdd wake up? one command (read/write) over sata/sas/ide channel wake it up? >> on linux raid we have a read algorithm and a write algorithm >> if a raid1 write occur all disks will wake up >> if a raid1 (raid0 or another) read occur only the disk will wake up >> >> but check you SATA/IDE/SATA controller, how it wake up your disk, and >> how you hdd wake up >> >> 2011/1/31 Piergiorgio Sartor <piergiorgio.sartor@nexgo.de>: >>> Hi all, >>> >>> assuming there is a NAS, with, for example, 10 HDDs >>> in RAID-6. Assuming the HDDs are put in standby, in >>> order to save energy, when the NAS is not used. >>> How is the spin up sequence when the corresponding >>> /dev/mdX device is accessed? >>> Will the system spin up one HDD at time or all together? >>> >>> Obviously, one at time will limit the peak current, >>> thus allowing a better dimensioned PSU, working almost >>> always around the optimal efficiency point. >>> >>> Thanks a lot for any information on the topic, >>> >>> bye, >>> >>> -- >>> >>> piergiorgio >>> -- >>> To unsubscribe from this list: send the line "unsubscribe linux-raid" in >>> the body of a message to majordomo@vger.kernel.org >>> More majordomo info at http://vger.kernel.org/majordomo-info.html >>> >> >> >> >> -- >> Roberto Spadim >> Spadim Technology / SPAEmpresarial >> -- >> To unsubscribe from this list: send the line "unsubscribe linux-raid" in >> the body of a message to majordomo@vger.kernel.org >> More majordomo info at http://vger.kernel.org/majordomo-info.html >> > > I would guess, that on a RAID0 setup, any read to that md device would > wake every disk up in that setup. No? > > // Mathias > > -- Roberto Spadim Spadim Technology / SPAEmpresarial -- To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html ^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: RAID HDDs spin up sequence 2011-01-31 21:25 ` Roberto Spadim @ 2011-01-31 21:29 ` Mathias Burén 2011-01-31 21:35 ` Roberto Spadim 0 siblings, 1 reply; 22+ messages in thread From: Mathias Burén @ 2011-01-31 21:29 UTC (permalink / raw) To: Roberto Spadim; +Cc: Piergiorgio Sartor, linux-raid On 31 January 2011 21:25, Roberto Spadim <roberto@spadim.com.br> wrote: > just if you need information on all disks > define your read. are you reading 1 byte or many bytes (on all hds?) > the number of bytes read/write is the point > if you need another disk to read/write you need to wake up your another hd > > check this implementations on mdadm: > linear, raid0, 0, stripe > maybe one is better for low power than another, but maybe one is > faster than another > > 2011/1/31 Mathias Burén <mathias.buren@gmail.com>: >> On 31 January 2011 21:09, Roberto Spadim <roberto@spadim.com.br> wrote: >>> you psu must be dimensioned to work with everythink at full work load >>> (it´s a real production NAS right?! not a test) >>> your SAS/IDE/SATA controller and HDD manual should be checked >>> how hdd wake up? one command (read/write) over sata/sas/ide channel wake it up? >>> on linux raid we have a read algorithm and a write algorithm >>> if a raid1 write occur all disks will wake up >>> if a raid1 (raid0 or another) read occur only the disk will wake up >>> >>> but check you SATA/IDE/SATA controller, how it wake up your disk, and >>> how you hdd wake up >>> >>> 2011/1/31 Piergiorgio Sartor <piergiorgio.sartor@nexgo.de>: >>>> Hi all, >>>> >>>> assuming there is a NAS, with, for example, 10 HDDs >>>> in RAID-6. Assuming the HDDs are put in standby, in >>>> order to save energy, when the NAS is not used. >>>> How is the spin up sequence when the corresponding >>>> /dev/mdX device is accessed? >>>> Will the system spin up one HDD at time or all together? >>>> >>>> Obviously, one at time will limit the peak current, >>>> thus allowing a better dimensioned PSU, working almost >>>> always around the optimal efficiency point. >>>> >>>> Thanks a lot for any information on the topic, >>>> >>>> bye, >>>> >>>> -- >>>> >>>> piergiorgio >>>> -- >>>> To unsubscribe from this list: send the line "unsubscribe linux-raid" in >>>> the body of a message to majordomo@vger.kernel.org >>>> More majordomo info at http://vger.kernel.org/majordomo-info.html >>>> >>> >>> >>> >>> -- >>> Roberto Spadim >>> Spadim Technology / SPAEmpresarial >>> -- >>> To unsubscribe from this list: send the line "unsubscribe linux-raid" in >>> the body of a message to majordomo@vger.kernel.org >>> More majordomo info at http://vger.kernel.org/majordomo-info.html >>> >> >> I would guess, that on a RAID0 setup, any read to that md device would >> wake every disk up in that setup. No? >> >> // Mathias >> >> > > > > -- > Roberto Spadim > Spadim Technology / SPAEmpresarial > I meant "default RAID0 setup with mdadm", with a minimum of 2 HDDs, and any fs on top of that. (with or without lvm) Unless the file is very small (smaller than the minimum piece of data that's being spread across all devices in the RAID0 setup (btw, that's this called in RAID0, is it chunk size? Stripe size?)) or cached, you'll wake the HDDs up. Right? // Mathias -- To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html ^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: RAID HDDs spin up sequence 2011-01-31 21:29 ` Mathias Burén @ 2011-01-31 21:35 ` Roberto Spadim 0 siblings, 0 replies; 22+ messages in thread From: Roberto Spadim @ 2011-01-31 21:35 UTC (permalink / raw) To: Mathias Burén; +Cc: Piergiorgio Sartor, linux-raid i think yes (hd wake up on write/read command) check linear and stripe layout linear is like a lvm (concatenation) i think raid0 work like stripe, but need to check it and return (raid0 = 0 = stripe level, linear!=raid0 for linux implementation) 2011/1/31 Mathias Burén <mathias.buren@gmail.com>: > On 31 January 2011 21:25, Roberto Spadim <roberto@spadim.com.br> wrote: >> just if you need information on all disks >> define your read. are you reading 1 byte or many bytes (on all hds?) >> the number of bytes read/write is the point >> if you need another disk to read/write you need to wake up your another hd >> >> check this implementations on mdadm: >> linear, raid0, 0, stripe >> maybe one is better for low power than another, but maybe one is >> faster than another >> >> 2011/1/31 Mathias Burén <mathias.buren@gmail.com>: >>> On 31 January 2011 21:09, Roberto Spadim <roberto@spadim.com.br> wrote: >>>> you psu must be dimensioned to work with everythink at full work load >>>> (it´s a real production NAS right?! not a test) >>>> your SAS/IDE/SATA controller and HDD manual should be checked >>>> how hdd wake up? one command (read/write) over sata/sas/ide channel wake it up? >>>> on linux raid we have a read algorithm and a write algorithm >>>> if a raid1 write occur all disks will wake up >>>> if a raid1 (raid0 or another) read occur only the disk will wake up >>>> >>>> but check you SATA/IDE/SATA controller, how it wake up your disk, and >>>> how you hdd wake up >>>> >>>> 2011/1/31 Piergiorgio Sartor <piergiorgio.sartor@nexgo.de>: >>>>> Hi all, >>>>> >>>>> assuming there is a NAS, with, for example, 10 HDDs >>>>> in RAID-6. Assuming the HDDs are put in standby, in >>>>> order to save energy, when the NAS is not used. >>>>> How is the spin up sequence when the corresponding >>>>> /dev/mdX device is accessed? >>>>> Will the system spin up one HDD at time or all together? >>>>> >>>>> Obviously, one at time will limit the peak current, >>>>> thus allowing a better dimensioned PSU, working almost >>>>> always around the optimal efficiency point. >>>>> >>>>> Thanks a lot for any information on the topic, >>>>> >>>>> bye, >>>>> >>>>> -- >>>>> >>>>> piergiorgio >>>>> -- >>>>> To unsubscribe from this list: send the line "unsubscribe linux-raid" in >>>>> the body of a message to majordomo@vger.kernel.org >>>>> More majordomo info at http://vger.kernel.org/majordomo-info.html >>>>> >>>> >>>> >>>> >>>> -- >>>> Roberto Spadim >>>> Spadim Technology / SPAEmpresarial >>>> -- >>>> To unsubscribe from this list: send the line "unsubscribe linux-raid" in >>>> the body of a message to majordomo@vger.kernel.org >>>> More majordomo info at http://vger.kernel.org/majordomo-info.html >>>> >>> >>> I would guess, that on a RAID0 setup, any read to that md device would >>> wake every disk up in that setup. No? >>> >>> // Mathias >>> >>> >> >> >> >> -- >> Roberto Spadim >> Spadim Technology / SPAEmpresarial >> > > I meant "default RAID0 setup with mdadm", with a minimum of 2 HDDs, > and any fs on top of that. (with or without lvm) > > Unless the file is very small (smaller than the minimum piece of data > that's being spread across all devices in the RAID0 setup (btw, that's > this called in RAID0, is it chunk size? Stripe size?)) or cached, > you'll wake the HDDs up. Right? > > // Mathias > -- > To unsubscribe from this list: send the line "unsubscribe linux-raid" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html > -- Roberto Spadim Spadim Technology / SPAEmpresarial -- To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html ^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: RAID HDDs spin up sequence 2011-01-31 21:09 ` Roberto Spadim 2011-01-31 21:10 ` Roberto Spadim 2011-01-31 21:11 ` Mathias Burén @ 2011-01-31 22:23 ` Piergiorgio Sartor 2011-01-31 22:42 ` Roberto Spadim 2011-02-01 13:55 ` brian.foster 2 siblings, 2 replies; 22+ messages in thread From: Piergiorgio Sartor @ 2011-01-31 22:23 UTC (permalink / raw) To: Roberto Spadim; +Cc: Piergiorgio Sartor, linux-raid On Mon, Jan 31, 2011 at 07:09:24PM -0200, Roberto Spadim wrote: > you psu must be dimensioned to work with everythink at full work load > (it´s a real production NAS right?! not a test) > your SAS/IDE/SATA controller and HDD manual should be checked > how hdd wake up? one command (read/write) over sata/sas/ide channel wake it up? > on linux raid we have a read algorithm and a write algorithm > if a raid1 write occur all disks will wake up > if a raid1 (raid0 or another) read occur only the disk will wake up > > but check you SATA/IDE/SATA controller, how it wake up your disk, and > how you hdd wake up Hi, thanks for the answer, unfortunately I was hoping to have made myself clear enough. First of all, it is a RAID-6, so let's say that's already decided by requirements. With SATA HDDs. Second, the question was exactly about how the HDDs are waked up. This is a SW issue, trying with normal setups, i.e. a couple of disks, it is possible to send them to sleep (hdparm -y /dev/hdX) and the wake them up by a simple access. I had no opportunity to check this with a RAID-5/6, so I was asking if anyone knows. Finally, in order to be power efficient, the PSU, assuming something like an 80 Plus Gold, should work at not less than 20% of the nominal power, otherwise (according to some reviews), the efficiency drops far below the 80%~90% declared by the 80 Plus standard (which is measured at 20%, 50% and 100% of the maximum specified power). It seem it gets easily around 40%~50%. So, the PSU must be somehow under dimensioned for the spin up of 10 HDDs, which seem to require a possible 30W*10=300W (some nasty HDDs seem to require 30W, in this situation) only for the storage. If the HDDs spin up one after the other, then the peak consumption is only 30W, which might allow a lower power PSU, in contrast with the requirement to provide 300W alone for the spin up. So, back to the original question, if a 10 HDDs RAID-6 is in standby, how do the single HDD will be waked up, in case of access? Of course, a quite larger access, i.e. some GiB of data. Thanks again, bye, -- piergiorgio -- To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html ^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: RAID HDDs spin up sequence 2011-01-31 22:23 ` Piergiorgio Sartor @ 2011-01-31 22:42 ` Roberto Spadim 2011-01-31 22:42 ` Roberto Spadim 2011-01-31 23:07 ` Piergiorgio Sartor 2011-02-01 13:55 ` brian.foster 1 sibling, 2 replies; 22+ messages in thread From: Roberto Spadim @ 2011-01-31 22:42 UTC (permalink / raw) To: Piergiorgio Sartor; +Cc: linux-raid the better probability here is: all disks must be waked up since you can have acces of 1gb but starting at a position that all disks must be used don´t try used small PSU HP Proliant ML310G5 start all hardware on power up (a lot o Watts) and after slow down thinks... why? check if PSU is ok, if not, don´t start server. that´s a good PSU system. ok if you want to test, i think the worst scenario is all disks beeing waked up, i think linux use async (many threads) commands to send write/read, maybe you will have a small time between wake up (maybe just some microseconds) 2011/1/31 Piergiorgio Sartor <piergiorgio.sartor@nexgo.de>: > On Mon, Jan 31, 2011 at 07:09:24PM -0200, Roberto Spadim wrote: >> you psu must be dimensioned to work with everythink at full work load >> (it愀 a real production NAS right?! not a test) >> your SAS/IDE/SATA controller and HDD manual should be checked >> how hdd wake up? one command (read/write) over sata/sas/ide channel wake it up? >> on linux raid we have a read algorithm and a write algorithm >> if a raid1 write occur all disks will wake up >> if a raid1 (raid0 or another) read occur only the disk will wake up >> >> but check you SATA/IDE/SATA controller, how it wake up your disk, and >> how you hdd wake up > > Hi, thanks for the answer, unfortunately I was > hoping to have made myself clear enough. > > First of all, it is a RAID-6, so let's say that's > already decided by requirements. With SATA HDDs. > > Second, the question was exactly about how the HDDs > are waked up. This is a SW issue, trying with normal > setups, i.e. a couple of disks, it is possible to > send them to sleep (hdparm -y /dev/hdX) and the wake > them up by a simple access. > I had no opportunity to check this with a RAID-5/6, > so I was asking if anyone knows. > > Finally, in order to be power efficient, the PSU, > assuming something like an 80 Plus Gold, should work > at not less than 20% of the nominal power, otherwise > (according to some reviews), the efficiency drops far > below the 80%~90% declared by the 80 Plus standard > (which is measured at 20%, 50% and 100% of the maximum > specified power). > It seem it gets easily around 40%~50%. > So, the PSU must be somehow under dimensioned for the > spin up of 10 HDDs, which seem to require a possible > 30W*10=300W (some nasty HDDs seem to require 30W, in > this situation) only for the storage. > > If the HDDs spin up one after the other, then the peak > consumption is only 30W, which might allow a lower > power PSU, in contrast with the requirement to provide > 300W alone for the spin up. > > So, back to the original question, if a 10 HDDs RAID-6 > is in standby, how do the single HDD will be waked up, > in case of access? Of course, a quite larger access, > i.e. some GiB of data. > > Thanks again, > > bye, > > -- > > piergiorgio > -- > To unsubscribe from this list: send the line "unsubscribe linux-raid" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html > -- Roberto Spadim Spadim Technology / SPAEmpresarial -- To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html ^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: RAID HDDs spin up sequence 2011-01-31 22:42 ` Roberto Spadim @ 2011-01-31 22:42 ` Roberto Spadim 2011-01-31 23:07 ` Piergiorgio Sartor 1 sibling, 0 replies; 22+ messages in thread From: Roberto Spadim @ 2011-01-31 22:42 UTC (permalink / raw) To: Piergiorgio Sartor; +Cc: linux-raid source code is the best source of information try here too (WIKI) https://raid.wiki.kernel.org/index.php/Main_Page 2011/1/31 Roberto Spadim <roberto@spadim.com.br>: > the better probability here is: all disks must be waked up > since you can have acces of 1gb but starting at a position that all > disks must be used > don´t try used small PSU > HP Proliant ML310G5 start all hardware on power up (a lot o Watts) and > after slow down thinks... why? check if PSU is ok, if not, don´t start > server. that´s a good PSU system. > ok if you want to test, i think the worst scenario is all disks beeing > waked up, i think linux use async (many threads) commands to send > write/read, maybe you will have a small time between wake up (maybe > just some microseconds) > > 2011/1/31 Piergiorgio Sartor <piergiorgio.sartor@nexgo.de>: >> On Mon, Jan 31, 2011 at 07:09:24PM -0200, Roberto Spadim wrote: >>> you psu must be dimensioned to work with everythink at full work load >>> (it愀 a real production NAS right?! not a test) >>> your SAS/IDE/SATA controller and HDD manual should be checked >>> how hdd wake up? one command (read/write) over sata/sas/ide channel wake it up? >>> on linux raid we have a read algorithm and a write algorithm >>> if a raid1 write occur all disks will wake up >>> if a raid1 (raid0 or another) read occur only the disk will wake up >>> >>> but check you SATA/IDE/SATA controller, how it wake up your disk, and >>> how you hdd wake up >> >> Hi, thanks for the answer, unfortunately I was >> hoping to have made myself clear enough. >> >> First of all, it is a RAID-6, so let's say that's >> already decided by requirements. With SATA HDDs. >> >> Second, the question was exactly about how the HDDs >> are waked up. This is a SW issue, trying with normal >> setups, i.e. a couple of disks, it is possible to >> send them to sleep (hdparm -y /dev/hdX) and the wake >> them up by a simple access. >> I had no opportunity to check this with a RAID-5/6, >> so I was asking if anyone knows. >> >> Finally, in order to be power efficient, the PSU, >> assuming something like an 80 Plus Gold, should work >> at not less than 20% of the nominal power, otherwise >> (according to some reviews), the efficiency drops far >> below the 80%~90% declared by the 80 Plus standard >> (which is measured at 20%, 50% and 100% of the maximum >> specified power). >> It seem it gets easily around 40%~50%. >> So, the PSU must be somehow under dimensioned for the >> spin up of 10 HDDs, which seem to require a possible >> 30W*10=300W (some nasty HDDs seem to require 30W, in >> this situation) only for the storage. >> >> If the HDDs spin up one after the other, then the peak >> consumption is only 30W, which might allow a lower >> power PSU, in contrast with the requirement to provide >> 300W alone for the spin up. >> >> So, back to the original question, if a 10 HDDs RAID-6 >> is in standby, how do the single HDD will be waked up, >> in case of access? Of course, a quite larger access, >> i.e. some GiB of data. >> >> Thanks again, >> >> bye, >> >> -- >> >> piergiorgio >> -- >> To unsubscribe from this list: send the line "unsubscribe linux-raid" in >> the body of a message to majordomo@vger.kernel.org >> More majordomo info at http://vger.kernel.org/majordomo-info.html >> > > > > -- > Roberto Spadim > Spadim Technology / SPAEmpresarial > -- Roberto Spadim Spadim Technology / SPAEmpresarial -- To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html ^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: RAID HDDs spin up sequence 2011-01-31 22:42 ` Roberto Spadim 2011-01-31 22:42 ` Roberto Spadim @ 2011-01-31 23:07 ` Piergiorgio Sartor 2011-01-31 23:12 ` Roberto Spadim 2011-02-01 1:45 ` Phillip Susi 1 sibling, 2 replies; 22+ messages in thread From: Piergiorgio Sartor @ 2011-01-31 23:07 UTC (permalink / raw) To: Roberto Spadim; +Cc: Piergiorgio Sartor, linux-raid On Mon, Jan 31, 2011 at 08:42:25PM -0200, Roberto Spadim wrote: > the better probability here is: all disks must be waked up > since you can have acces of 1gb but starting at a position that all > disks must be used > don´t try used small PSU No, that's not an option, HDDs must have staggered spin up, which is, to certain extent, supported by some SATA HDD manufactures (not all, it seems). > HP Proliant ML310G5 start all hardware on power up (a lot o Watts) and > after slow down thinks... why? check if PSU is ok, if not, don´t start > server. that´s a good PSU system. As I wrote, efficiency is paramount here. > ok if you want to test, i think the worst scenario is all disks beeing > waked up, i think linux use async (many threads) commands to send > write/read, maybe you will have a small time between wake up (maybe > just some microseconds) That's interesting information. I suspected something like async access could have happened. Is this confirmed? Are the requests from md to the device layer asynchronous? How about libata? Does this serialize wake ups? Thanks, bye, pg > 2011/1/31 Piergiorgio Sartor <piergiorgio.sartor@nexgo.de>: > > On Mon, Jan 31, 2011 at 07:09:24PM -0200, Roberto Spadim wrote: > >> you psu must be dimensioned to work with everythink at full work load > >> (it愀 a real production NAS right?! not a test) > >> your SAS/IDE/SATA controller and HDD manual should be checked > >> how hdd wake up? one command (read/write) over sata/sas/ide channel wake it up? > >> on linux raid we have a read algorithm and a write algorithm > >> if a raid1 write occur all disks will wake up > >> if a raid1 (raid0 or another) read occur only the disk will wake up > >> > >> but check you SATA/IDE/SATA controller, how it wake up your disk, and > >> how you hdd wake up > > > > Hi, thanks for the answer, unfortunately I was > > hoping to have made myself clear enough. > > > > First of all, it is a RAID-6, so let's say that's > > already decided by requirements. With SATA HDDs. > > > > Second, the question was exactly about how the HDDs > > are waked up. This is a SW issue, trying with normal > > setups, i.e. a couple of disks, it is possible to > > send them to sleep (hdparm -y /dev/hdX) and the wake > > them up by a simple access. > > I had no opportunity to check this with a RAID-5/6, > > so I was asking if anyone knows. > > > > Finally, in order to be power efficient, the PSU, > > assuming something like an 80 Plus Gold, should work > > at not less than 20% of the nominal power, otherwise > > (according to some reviews), the efficiency drops far > > below the 80%~90% declared by the 80 Plus standard > > (which is measured at 20%, 50% and 100% of the maximum > > specified power). > > It seem it gets easily around 40%~50%. > > So, the PSU must be somehow under dimensioned for the > > spin up of 10 HDDs, which seem to require a possible > > 30W*10=300W (some nasty HDDs seem to require 30W, in > > this situation) only for the storage. > > > > If the HDDs spin up one after the other, then the peak > > consumption is only 30W, which might allow a lower > > power PSU, in contrast with the requirement to provide > > 300W alone for the spin up. > > > > So, back to the original question, if a 10 HDDs RAID-6 > > is in standby, how do the single HDD will be waked up, > > in case of access? Of course, a quite larger access, > > i.e. some GiB of data. > > > > Thanks again, > > > > bye, > > > > -- > > > > piergiorgio > > -- > > To unsubscribe from this list: send the line "unsubscribe linux-raid" in > > the body of a message to majordomo@vger.kernel.org > > More majordomo info at http://vger.kernel.org/majordomo-info.html > > > > > > -- > Roberto Spadim > Spadim Technology / SPAEmpresarial -- piergiorgio -- To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html ^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: RAID HDDs spin up sequence 2011-01-31 23:07 ` Piergiorgio Sartor @ 2011-01-31 23:12 ` Roberto Spadim 2011-02-01 1:45 ` Phillip Susi 1 sibling, 0 replies; 22+ messages in thread From: Roberto Spadim @ 2011-01-31 23:12 UTC (permalink / raw) To: Piergiorgio Sartor; +Cc: linux-raid you have scheduler (elevator) maybe checking this code is nice, but it´s device based (linux) check /sys/block/xxxx/ to make some tweaks i don´t know the real implementation (must check source code) but i think mdadm is over linux device layer (it don´t controll access to disk), i think it´s async, and linux kernel make thinks more sync... check you raid level source code somethink like: raid6.c it´s small i think it´s async does you know where linux kernel can get information (inside /sys or /proc filesystem) to know if you disk is weak? maybe a check and wait before send read command can be implemented... 2011/1/31 Piergiorgio Sartor <piergiorgio.sartor@nexgo.de>: > On Mon, Jan 31, 2011 at 08:42:25PM -0200, Roberto Spadim wrote: >> the better probability here is: all disks must be waked up >> since you can have acces of 1gb but starting at a position that all >> disks must be used >> don´t try used small PSU > > No, that's not an option, HDDs must have staggered > spin up, which is, to certain extent, supported by > some SATA HDD manufactures (not all, it seems). > >> HP Proliant ML310G5 start all hardware on power up (a lot o Watts) and >> after slow down thinks... why? check if PSU is ok, if not, don´t start >> server. that´s a good PSU system. > > As I wrote, efficiency is paramount here. > >> ok if you want to test, i think the worst scenario is all disks beeing >> waked up, i think linux use async (many threads) commands to send >> write/read, maybe you will have a small time between wake up (maybe >> just some microseconds) > > That's interesting information. I suspected something > like async access could have happened. > Is this confirmed? Are the requests from md to the > device layer asynchronous? > How about libata? Does this serialize wake ups? > > Thanks, > > bye, > > pg > >> 2011/1/31 Piergiorgio Sartor <piergiorgio.sartor@nexgo.de>: >> > On Mon, Jan 31, 2011 at 07:09:24PM -0200, Roberto Spadim wrote: >> >> you psu must be dimensioned to work with everythink at full work load >> >> (it愀 a real production NAS right?! not a test) >> >> your SAS/IDE/SATA controller and HDD manual should be checked >> >> how hdd wake up? one command (read/write) over sata/sas/ide channel wake it up? >> >> on linux raid we have a read algorithm and a write algorithm >> >> if a raid1 write occur all disks will wake up >> >> if a raid1 (raid0 or another) read occur only the disk will wake up >> >> >> >> but check you SATA/IDE/SATA controller, how it wake up your disk, and >> >> how you hdd wake up >> > >> > Hi, thanks for the answer, unfortunately I was >> > hoping to have made myself clear enough. >> > >> > First of all, it is a RAID-6, so let's say that's >> > already decided by requirements. With SATA HDDs. >> > >> > Second, the question was exactly about how the HDDs >> > are waked up. This is a SW issue, trying with normal >> > setups, i.e. a couple of disks, it is possible to >> > send them to sleep (hdparm -y /dev/hdX) and the wake >> > them up by a simple access. >> > I had no opportunity to check this with a RAID-5/6, >> > so I was asking if anyone knows. >> > >> > Finally, in order to be power efficient, the PSU, >> > assuming something like an 80 Plus Gold, should work >> > at not less than 20% of the nominal power, otherwise >> > (according to some reviews), the efficiency drops far >> > below the 80%~90% declared by the 80 Plus standard >> > (which is measured at 20%, 50% and 100% of the maximum >> > specified power). >> > It seem it gets easily around 40%~50%. >> > So, the PSU must be somehow under dimensioned for the >> > spin up of 10 HDDs, which seem to require a possible >> > 30W*10=300W (some nasty HDDs seem to require 30W, in >> > this situation) only for the storage. >> > >> > If the HDDs spin up one after the other, then the peak >> > consumption is only 30W, which might allow a lower >> > power PSU, in contrast with the requirement to provide >> > 300W alone for the spin up. >> > >> > So, back to the original question, if a 10 HDDs RAID-6 >> > is in standby, how do the single HDD will be waked up, >> > in case of access? Of course, a quite larger access, >> > i.e. some GiB of data. >> > >> > Thanks again, >> > >> > bye, >> > >> > -- >> > >> > piergiorgio >> > -- >> > To unsubscribe from this list: send the line "unsubscribe linux-raid" in >> > the body of a message to majordomo@vger.kernel.org >> > More majordomo info at http://vger.kernel.org/majordomo-info.html >> > >> >> >> >> -- >> Roberto Spadim >> Spadim Technology / SPAEmpresarial > > -- > > piergiorgio > -- > To unsubscribe from this list: send the line "unsubscribe linux-raid" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html > -- Roberto Spadim Spadim Technology / SPAEmpresarial -- To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html ^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: RAID HDDs spin up sequence 2011-01-31 23:07 ` Piergiorgio Sartor 2011-01-31 23:12 ` Roberto Spadim @ 2011-02-01 1:45 ` Phillip Susi 2011-02-01 12:39 ` Roman Mamedov 2011-02-01 13:10 ` John Robinson 1 sibling, 2 replies; 22+ messages in thread From: Phillip Susi @ 2011-02-01 1:45 UTC (permalink / raw) To: Piergiorgio Sartor; +Cc: Roberto Spadim, linux-raid On 01/31/2011 06:07 PM, Piergiorgio Sartor wrote: > That's interesting information. I suspected something > like async access could have happened. > Is this confirmed? Are the requests from md to the > device layer asynchronous? Yes, it sends down requests to all drives simultaneously to achieve maximum performance. This is the main purpose of raid-5. > How about libata? Does this serialize wake ups? It doesn't. It will wake them all up at once. If you want to force serial wakeup, you would need to do some hacking in libata or the scsi layer to only allow one wakeup request at once. Of course, with a 6 disk raid where each disk takes 6 seconds to spin up, you will be waiting 36 seconds to get your data. Not exactly desirable. ^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: RAID HDDs spin up sequence 2011-02-01 1:45 ` Phillip Susi @ 2011-02-01 12:39 ` Roman Mamedov 2011-02-01 13:10 ` John Robinson 1 sibling, 0 replies; 22+ messages in thread From: Roman Mamedov @ 2011-02-01 12:39 UTC (permalink / raw) To: Phillip Susi; +Cc: Piergiorgio Sartor, Roberto Spadim, linux-raid [-- Attachment #1: Type: text/plain, Size: 715 bytes --] On Mon, 31 Jan 2011 20:45:22 -0500 Phillip Susi <psusi@cfl.rr.com> wrote: > If you want to force serial wakeup, you would need to do some hacking in > libata or the scsi layer to only allow one wakeup request at once. Of > course, with a 6 disk raid where each disk takes 6 seconds to spin up, > you will be waiting 36 seconds to get your data. Not exactly desirable. Well, if the other option is to have your system hard-reset because the PSU overloads from 12-20-24 disks spinning up at once (you know staggered spin-up is supported by disks/controllers for a reason), then the wait does not look so bad. My take on the problem: just don't spin down your RAIDs. -- With respect, Roman [-- Attachment #2: signature.asc --] [-- Type: application/pgp-signature, Size: 198 bytes --] ^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: RAID HDDs spin up sequence 2011-02-01 1:45 ` Phillip Susi 2011-02-01 12:39 ` Roman Mamedov @ 2011-02-01 13:10 ` John Robinson 2011-02-01 21:37 ` Stan Hoeppner 1 sibling, 1 reply; 22+ messages in thread From: John Robinson @ 2011-02-01 13:10 UTC (permalink / raw) To: Phillip Susi; +Cc: Piergiorgio Sartor, linux-raid On 01/02/2011 01:45, Phillip Susi wrote: > On 01/31/2011 06:07 PM, Piergiorgio Sartor wrote: >> That's interesting information. I suspected something >> like async access could have happened. >> Is this confirmed? Are the requests from md to the >> device layer asynchronous? > > Yes, it sends down requests to all drives simultaneously to achieve > maximum performance. This is the main purpose of raid-5. > >> How about libata? Does this serialize wake ups? > > It doesn't. It will wake them all up at once. > > If you want to force serial wakeup, you would need to do some hacking in > libata or the scsi layer to only allow one wakeup request at once. Of > course, with a 6 disk raid where each disk takes 6 seconds to spin up, > you will be waiting 36 seconds to get your data. Not exactly desirable. That'd be an excessive amount of time to wait. A quarter of a second is more than enough, a tenth of a second would probably be enough. It's just the motor inrush current you're trying to avoid having simultaneously. So waiting another second for your array to wake up would mean you could use a sensibly-sized PSU operating in its 80%+ efficiency range, rather than a huge PSU operating inefficiently. Cheers, John. ^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: RAID HDDs spin up sequence 2011-02-01 13:10 ` John Robinson @ 2011-02-01 21:37 ` Stan Hoeppner 2011-02-01 22:46 ` Roberto Spadim 0 siblings, 1 reply; 22+ messages in thread From: Stan Hoeppner @ 2011-02-01 21:37 UTC (permalink / raw) To: John Robinson; +Cc: Phillip Susi, Piergiorgio Sartor, linux-raid John Robinson put forth on 2/1/2011 7:10 AM: > That'd be an excessive amount of time to wait. A quarter of a second is more > than enough, a tenth of a second would probably be enough. It's just the motor > inrush current you're trying to avoid having simultaneously. The blowers in a typical 2U server chassis will have slightly more startup current draw than the drives, assuming 5 80mm blowers and 8 2.5" drives. Mobos don't do staggered startup of blowers. Thus, staggering the drive spin up is pointless. Add to that the fact that most server chassis ship with PSUs large enough to carry the current draw of anything/everything you can stuff into them. > So waiting another second for your array to wake up would mean you could use a > sensibly-sized PSU operating in its 80%+ efficiency range, rather than a huge > PSU operating inefficiently. A typical 2.5" 10K RPM 600GB enterprise HDD, such as the Seagate Savvio, has a startup draw of 24.1 watts combined from the 12v and 5v rails. A RAID/JBOD chassis of 24 such drives, which is sold by dozens of vendors today, will draw only 578.4 watts with all drives spinning up concurrently. Most such chassis on the market today are sold with 800w to 1800w redundant PSUs, again, making staggered spin up moot. -- Stan ^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: RAID HDDs spin up sequence 2011-02-01 21:37 ` Stan Hoeppner @ 2011-02-01 22:46 ` Roberto Spadim 0 siblings, 0 replies; 22+ messages in thread From: Roberto Spadim @ 2011-02-01 22:46 UTC (permalink / raw) To: Stan Hoeppner; +Cc: John Robinson, Phillip Susi, Piergiorgio Sartor, linux-raid some hardware solutions.... change PSU change hard disk to solid state disks (OCZ is fast, energy saver, with high MTBF, a very good disk is OCZ VERTEX 2, REVODRIVE (pci-express) and OCZ AGILITY 2, check read random 4kb >=200 MB/s is better than any SAS disk... latency is <0.1ms) maybe changing to solid state is a good solution (cost is the point here) another point is when startup remove (hot plug) some disks, and plug after 2011/2/1 Stan Hoeppner <stan@hardwarefreak.com>: > John Robinson put forth on 2/1/2011 7:10 AM: > >> That'd be an excessive amount of time to wait. A quarter of a second is more >> than enough, a tenth of a second would probably be enough. It's just the motor >> inrush current you're trying to avoid having simultaneously. > > The blowers in a typical 2U server chassis will have slightly more startup > current draw than the drives, assuming 5 80mm blowers and 8 2.5" drives. Mobos > don't do staggered startup of blowers. Thus, staggering the drive spin up is > pointless. Add to that the fact that most server chassis ship with PSUs large > enough to carry the current draw of anything/everything you can stuff into them. > >> So waiting another second for your array to wake up would mean you could use a >> sensibly-sized PSU operating in its 80%+ efficiency range, rather than a huge >> PSU operating inefficiently. > > A typical 2.5" 10K RPM 600GB enterprise HDD, such as the Seagate Savvio, has a > startup draw of 24.1 watts combined from the 12v and 5v rails. A RAID/JBOD > chassis of 24 such drives, which is sold by dozens of vendors today, will draw > only 578.4 watts with all drives spinning up concurrently. Most such chassis on > the market today are sold with 800w to 1800w redundant PSUs, again, making > staggered spin up moot. > > -- > Stan > -- > To unsubscribe from this list: send the line "unsubscribe linux-raid" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html > -- Roberto Spadim Spadim Technology / SPAEmpresarial -- To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html ^ permalink raw reply [flat|nested] 22+ messages in thread
* RE: RAID HDDs spin up sequence 2011-01-31 22:23 ` Piergiorgio Sartor 2011-01-31 22:42 ` Roberto Spadim @ 2011-02-01 13:55 ` brian.foster 2011-02-01 14:37 ` Roberto Spadim 1 sibling, 1 reply; 22+ messages in thread From: brian.foster @ 2011-02-01 13:55 UTC (permalink / raw) To: piergiorgio.sartor, roberto; +Cc: linux-raid > -----Original Message----- > From: linux-raid-owner@vger.kernel.org [mailto:linux-raid- > owner@vger.kernel.org] On Behalf Of Piergiorgio Sartor > Sent: Monday, January 31, 2011 5:24 PM > To: Roberto Spadim > Cc: Piergiorgio Sartor; linux-raid@vger.kernel.org > Subject: Re: RAID HDDs spin up sequence > > On Mon, Jan 31, 2011 at 07:09:24PM -0200, Roberto Spadim wrote: > > you psu must be dimensioned to work with everythink at full work load > > (it´s a real production NAS right?! not a test) > > your SAS/IDE/SATA controller and HDD manual should be checked > > how hdd wake up? one command (read/write) over sata/sas/ide channel > wake it up? > > on linux raid we have a read algorithm and a write algorithm > > if a raid1 write occur all disks will wake up > > if a raid1 (raid0 or another) read occur only the disk will wake up > > > > but check you SATA/IDE/SATA controller, how it wake up your disk, and > > how you hdd wake up > > Hi, thanks for the answer, unfortunately I was > hoping to have made myself clear enough. > > First of all, it is a RAID-6, so let's say that's > already decided by requirements. With SATA HDDs. > > Second, the question was exactly about how the HDDs > are waked up. This is a SW issue, trying with normal > setups, i.e. a couple of disks, it is possible to > send them to sleep (hdparm -y /dev/hdX) and the wake > them up by a simple access. > I had no opportunity to check this with a RAID-5/6, > so I was asking if anyone knows. > > Finally, in order to be power efficient, the PSU, > assuming something like an 80 Plus Gold, should work > at not less than 20% of the nominal power, otherwise > (according to some reviews), the efficiency drops far > below the 80%~90% declared by the 80 Plus standard > (which is measured at 20%, 50% and 100% of the maximum > specified power). > It seem it gets easily around 40%~50%. > So, the PSU must be somehow under dimensioned for the > spin up of 10 HDDs, which seem to require a possible > 30W*10=300W (some nasty HDDs seem to require 30W, in > this situation) only for the storage. > > If the HDDs spin up one after the other, then the peak > consumption is only 30W, which might allow a lower > power PSU, in contrast with the requirement to provide > 300W alone for the spin up. > > So, back to the original question, if a 10 HDDs RAID-6 > is in standby, how do the single HDD will be waked up, > in case of access? Of course, a quite larger access, > i.e. some GiB of data. We have a similar situation with our Iomega NAS products. We had a fairly crude locking mechanism implemented at the SCSI level that suits our needs to support staggered spin up. As indicated previously, we find that a 1 or 2 second delay between spin ups modulates the current draw enough such that we don't run into problems. We use this in conjunction with MD without much of a problem. I'm not sure that the code as implemented is appropriate for mainline inclusion (and I'm not going to post it directly), but FYI the patches are included in the open source tarball that is made available on the support section of the iomega.com website (hint: check out the ix12 support & downloads section). Brian > > Thanks again, > > bye, > > -- > > piergiorgio > -- > To unsubscribe from this list: send the line "unsubscribe linux-raid" > in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html -- To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html ^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: RAID HDDs spin up sequence 2011-02-01 13:55 ` brian.foster @ 2011-02-01 14:37 ` Roberto Spadim 2011-02-01 14:44 ` Roberto Spadim 0 siblings, 1 reply; 22+ messages in thread From: Roberto Spadim @ 2011-02-01 14:37 UTC (permalink / raw) To: brian.foster; +Cc: piergiorgio.sartor, linux-raid maybe a check inside mdadm code should be implemented before send read/write command check if device is sleeping, if yes if previous devices wakeup command was more than x seconds OR no device, send a wakeup command another idea: a fixed delay to wake up, diferent per device, but could only work when we send(before sleep part of code) to all devices command at the same time mdadm should be changed (options be included) and kernel source too 2011/2/1 <brian.foster@emc.com>: >> -----Original Message----- >> From: linux-raid-owner@vger.kernel.org [mailto:linux-raid- >> owner@vger.kernel.org] On Behalf Of Piergiorgio Sartor >> Sent: Monday, January 31, 2011 5:24 PM >> To: Roberto Spadim >> Cc: Piergiorgio Sartor; linux-raid@vger.kernel.org >> Subject: Re: RAID HDDs spin up sequence >> >> On Mon, Jan 31, 2011 at 07:09:24PM -0200, Roberto Spadim wrote: >> > you psu must be dimensioned to work with everythink at full work load >> > (it愀 a real production NAS right?! not a test) >> > your SAS/IDE/SATA controller and HDD manual should be checked >> > how hdd wake up? one command (read/write) over sata/sas/ide channel >> wake it up? >> > on linux raid we have a read algorithm and a write algorithm >> > if a raid1 write occur all disks will wake up >> > if a raid1 (raid0 or another) read occur only the disk will wake up >> > >> > but check you SATA/IDE/SATA controller, how it wake up your disk, and >> > how you hdd wake up >> >> Hi, thanks for the answer, unfortunately I was >> hoping to have made myself clear enough. >> >> First of all, it is a RAID-6, so let's say that's >> already decided by requirements. With SATA HDDs. >> >> Second, the question was exactly about how the HDDs >> are waked up. This is a SW issue, trying with normal >> setups, i.e. a couple of disks, it is possible to >> send them to sleep (hdparm -y /dev/hdX) and the wake >> them up by a simple access. >> I had no opportunity to check this with a RAID-5/6, >> so I was asking if anyone knows. >> >> Finally, in order to be power efficient, the PSU, >> assuming something like an 80 Plus Gold, should work >> at not less than 20% of the nominal power, otherwise >> (according to some reviews), the efficiency drops far >> below the 80%~90% declared by the 80 Plus standard >> (which is measured at 20%, 50% and 100% of the maximum >> specified power). >> It seem it gets easily around 40%~50%. >> So, the PSU must be somehow under dimensioned for the >> spin up of 10 HDDs, which seem to require a possible >> 30W*10=300W (some nasty HDDs seem to require 30W, in >> this situation) only for the storage. >> >> If the HDDs spin up one after the other, then the peak >> consumption is only 30W, which might allow a lower >> power PSU, in contrast with the requirement to provide >> 300W alone for the spin up. >> >> So, back to the original question, if a 10 HDDs RAID-6 >> is in standby, how do the single HDD will be waked up, >> in case of access? Of course, a quite larger access, >> i.e. some GiB of data. > > We have a similar situation with our Iomega NAS products. We had a fairly crude locking mechanism implemented at the SCSI level that suits our needs to support staggered spin up. As indicated previously, we find that a 1 or 2 second delay between spin ups modulates the current draw enough such that we don't run into problems. We use this in conjunction with MD without much of a problem. > > I'm not sure that the code as implemented is appropriate for mainline inclusion (and I'm not going to post it directly), but FYI the patches are included in the open source tarball that is made available on the support section of the iomega.com website (hint: check out the ix12 support & downloads section). > > Brian > >> >> Thanks again, >> >> bye, >> >> -- >> >> piergiorgio >> -- >> To unsubscribe from this list: send the line "unsubscribe linux-raid" >> in >> the body of a message to majordomo@vger.kernel.org >> More majordomo info at http://vger.kernel.org/majordomo-info.html > > -- > To unsubscribe from this list: send the line "unsubscribe linux-raid" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html > -- Roberto Spadim Spadim Technology / SPAEmpresarial -- To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html ^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: RAID HDDs spin up sequence 2011-02-01 14:37 ` Roberto Spadim @ 2011-02-01 14:44 ` Roberto Spadim 2011-02-01 15:01 ` John Robinson 0 siblings, 1 reply; 22+ messages in thread From: Roberto Spadim @ 2011-02-01 14:44 UTC (permalink / raw) To: brian.foster; +Cc: piergiorgio.sartor, linux-raid try to put big capacitors and inductors on energy cable... it´s a good eletric level solution... capacitor and inductors will help the start current/voltage oscilation 2011/2/1 Roberto Spadim <roberto@spadim.com.br>: > maybe a check inside mdadm code should be implemented > > before send read/write command check if device is sleeping, > if yes > if previous devices wakeup command was more than x seconds OR no > device, send a wakeup command > > another idea: a fixed delay to wake up, diferent per device, but could > only work when we send(before sleep part of code) to all devices > command at the same time > > mdadm should be changed (options be included) and kernel source too > > 2011/2/1 <brian.foster@emc.com>: >>> -----Original Message----- >>> From: linux-raid-owner@vger.kernel.org [mailto:linux-raid- >>> owner@vger.kernel.org] On Behalf Of Piergiorgio Sartor >>> Sent: Monday, January 31, 2011 5:24 PM >>> To: Roberto Spadim >>> Cc: Piergiorgio Sartor; linux-raid@vger.kernel.org >>> Subject: Re: RAID HDDs spin up sequence >>> >>> On Mon, Jan 31, 2011 at 07:09:24PM -0200, Roberto Spadim wrote: >>> > you psu must be dimensioned to work with everythink at full work load >>> > (it愀 a real production NAS right?! not a test) >>> > your SAS/IDE/SATA controller and HDD manual should be checked >>> > how hdd wake up? one command (read/write) over sata/sas/ide channel >>> wake it up? >>> > on linux raid we have a read algorithm and a write algorithm >>> > if a raid1 write occur all disks will wake up >>> > if a raid1 (raid0 or another) read occur only the disk will wake up >>> > >>> > but check you SATA/IDE/SATA controller, how it wake up your disk, and >>> > how you hdd wake up >>> >>> Hi, thanks for the answer, unfortunately I was >>> hoping to have made myself clear enough. >>> >>> First of all, it is a RAID-6, so let's say that's >>> already decided by requirements. With SATA HDDs. >>> >>> Second, the question was exactly about how the HDDs >>> are waked up. This is a SW issue, trying with normal >>> setups, i.e. a couple of disks, it is possible to >>> send them to sleep (hdparm -y /dev/hdX) and the wake >>> them up by a simple access. >>> I had no opportunity to check this with a RAID-5/6, >>> so I was asking if anyone knows. >>> >>> Finally, in order to be power efficient, the PSU, >>> assuming something like an 80 Plus Gold, should work >>> at not less than 20% of the nominal power, otherwise >>> (according to some reviews), the efficiency drops far >>> below the 80%~90% declared by the 80 Plus standard >>> (which is measured at 20%, 50% and 100% of the maximum >>> specified power). >>> It seem it gets easily around 40%~50%. >>> So, the PSU must be somehow under dimensioned for the >>> spin up of 10 HDDs, which seem to require a possible >>> 30W*10=300W (some nasty HDDs seem to require 30W, in >>> this situation) only for the storage. >>> >>> If the HDDs spin up one after the other, then the peak >>> consumption is only 30W, which might allow a lower >>> power PSU, in contrast with the requirement to provide >>> 300W alone for the spin up. >>> >>> So, back to the original question, if a 10 HDDs RAID-6 >>> is in standby, how do the single HDD will be waked up, >>> in case of access? Of course, a quite larger access, >>> i.e. some GiB of data. >> >> We have a similar situation with our Iomega NAS products. We had a fairly crude locking mechanism implemented at the SCSI level that suits our needs to support staggered spin up. As indicated previously, we find that a 1 or 2 second delay between spin ups modulates the current draw enough such that we don't run into problems. We use this in conjunction with MD without much of a problem. >> >> I'm not sure that the code as implemented is appropriate for mainline inclusion (and I'm not going to post it directly), but FYI the patches are included in the open source tarball that is made available on the support section of the iomega.com website (hint: check out the ix12 support & downloads section). >> >> Brian >> >>> >>> Thanks again, >>> >>> bye, >>> >>> -- >>> >>> piergiorgio >>> -- >>> To unsubscribe from this list: send the line "unsubscribe linux-raid" >>> in >>> the body of a message to majordomo@vger.kernel.org >>> More majordomo info at http://vger.kernel.org/majordomo-info.html >> >> -- >> To unsubscribe from this list: send the line "unsubscribe linux-raid" in >> the body of a message to majordomo@vger.kernel.org >> More majordomo info at http://vger.kernel.org/majordomo-info.html >> > > > > -- > Roberto Spadim > Spadim Technology / SPAEmpresarial > -- Roberto Spadim Spadim Technology / SPAEmpresarial -- To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html ^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: RAID HDDs spin up sequence 2011-02-01 14:44 ` Roberto Spadim @ 2011-02-01 15:01 ` John Robinson 2011-02-01 15:46 ` Roberto Spadim 0 siblings, 1 reply; 22+ messages in thread From: John Robinson @ 2011-02-01 15:01 UTC (permalink / raw) To: Roberto Spadim; +Cc: linux-raid On 01/02/2011 14:44, Roberto Spadim wrote: > try to put big capacitors and inductors on energy cable... it´s a good > eletric level solution... capacitor and inductors will help the start > current/voltage oscilation No, it isn't; they'd cause excessive power draw on initial power-up. Cheers, John. -- To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html ^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: RAID HDDs spin up sequence 2011-02-01 15:01 ` John Robinson @ 2011-02-01 15:46 ` Roberto Spadim 0 siblings, 0 replies; 22+ messages in thread From: Roberto Spadim @ 2011-02-01 15:46 UTC (permalink / raw) To: John Robinson; +Cc: linux-raid hum that´s right. maybe a power-up safe eletronic could be implemented i don´t know if implementation is faster out of linux (hardware) or inside (code) using code is easy to port it to any hardware maybe neil have interest in implement it in source code (i don´t know if there´s more linux (kernel) developer here in list) 2011/2/1 John Robinson <john.robinson@anonymous.org.uk>: > On 01/02/2011 14:44, Roberto Spadim wrote: >> >> try to put big capacitors and inductors on energy cable... it´s a good >> eletric level solution... capacitor and inductors will help the start >> current/voltage oscilation > > No, it isn't; they'd cause excessive power draw on initial power-up. > > Cheers, > > John. > > -- > To unsubscribe from this list: send the line "unsubscribe linux-raid" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html > -- Roberto Spadim Spadim Technology / SPAEmpresarial -- To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html ^ permalink raw reply [flat|nested] 22+ messages in thread
end of thread, other threads:[~2011-02-01 22:46 UTC | newest] Thread overview: 22+ messages (download: mbox.gz / follow: Atom feed) -- links below jump to the message on this page -- 2011-01-31 20:18 RAID HDDs spin up sequence Piergiorgio Sartor 2011-01-31 21:09 ` Roberto Spadim 2011-01-31 21:10 ` Roberto Spadim 2011-01-31 21:11 ` Mathias Burén 2011-01-31 21:25 ` Roberto Spadim 2011-01-31 21:29 ` Mathias Burén 2011-01-31 21:35 ` Roberto Spadim 2011-01-31 22:23 ` Piergiorgio Sartor 2011-01-31 22:42 ` Roberto Spadim 2011-01-31 22:42 ` Roberto Spadim 2011-01-31 23:07 ` Piergiorgio Sartor 2011-01-31 23:12 ` Roberto Spadim 2011-02-01 1:45 ` Phillip Susi 2011-02-01 12:39 ` Roman Mamedov 2011-02-01 13:10 ` John Robinson 2011-02-01 21:37 ` Stan Hoeppner 2011-02-01 22:46 ` Roberto Spadim 2011-02-01 13:55 ` brian.foster 2011-02-01 14:37 ` Roberto Spadim 2011-02-01 14:44 ` Roberto Spadim 2011-02-01 15:01 ` John Robinson 2011-02-01 15:46 ` Roberto Spadim
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.