HP Z840: Z840, bunch of several Raid 1 s

Similar Questions

• Several RAID, same box arrays

  From my small medical practice with a server on site (PE320) with H710 PERC for RAID 1 arrays and RAID 10 planned.  The content of my questions exposes my ignorance.  I use the application lifecycle controller to try to configure things.  I had thought about putting MS Server 2012 matrix RAID 1 and plop all my other apps (based in the main MySQL) on RAID 10 side of things.  Is this possible?  It is even a good idea?  Or just take all 6 2.5 "15K 300 GB drives and put them in a RAID 10 array.  It's a small office with only 4-5 clients.  Doubt performance issues will be great and plan on backup by other means.  Oh, and that's okay designate individual tables (1 and 10) that VDs?  Looking for a lighting...

  Dr. P

  Joseph Perkinson

  It is possible and in almost every respect them the recommended configuration. The reason is that it allows to separate data from the OS. This adds a level of stability, as well as the protects the data of all the problems that may arise from the BONE. It has a performance slightly decline, passing from a Raid 10 straight, but it's negligible. With desktop size you described, it would not even noticeable.

  That's right, they are called VDs for virtual disks.

  Let me know if this can help answer your questions, or if you have any other.

• Reuse a RAID disk and hot spare procedure

  My questions are for a PE R720xd with the PERC RAID internal H710P controller.

  I have replaced several RAID drives in the past, but I never did this before, so I was hoping to confirm my procedure for hard drives hot swap. A drive failed on one of my live servers. In the past I would always bring back the server during a maintenance window and replace the drive by using the configuration utility to make sure everything went correctly. This particular server has become a system of very high availability, so I was hoping to use for hot spare option do without bring down the server.

  The other issue is that Dell is not in stock on the disk, I need so I intend to use a same car I replaced recently on a server off-production that has already been fixed in a RAID1 in another Dell R720 Server table. Any time, I have never replaced a RAID drive, I always used a new drive out-of-the-box, so I'm also not 100% on how to prepare the drive to use for a new RAID array.

  I thought I am brings the donor server down and come off the drive I want to enjoy this the RAID array server using the configuration utility. So a question is - is there anything else I need to do preparation used in car for use in the new table?

  With respect to hot-swapping which lead to the new array, is it really as simple as just popping one out and put a new in? Can I do to confirm it has taken - apart from just confirming that the lights blink like crazy during synchronization?

  Thanks for your understanding!

  C.Short,

  You can switch the drive, but it will be not just to rebuild there is data on the hard drive. So when this disc is inserted, it will show foreign. You will need to do is to boot to the controller (Ctrl + R) and then CLEAR the foreign Configuration. Once completed, then you will be able to rebuild the disk on the virtual disk, you can do the BIOS of the controller, or from the OpenManage web interface (depending on the operating system).

  Let me know if it helps.

• Levels of raid Volumes Compellent

  Hello.

  We are looking into buying a Bay Compellent, Dell, but before I look for in its technology.

  Unfortunately, the few public records on the SC8000 are very brief.

  I'm looking for information how the table handles its raids, from what ive understood, when you create a volume, some data will be placed on raid 10, some on raid 5 and some might end up on raid6 (prioritization).

  But the definition of these levels of raid? They simply come from all disks of a certain type, say raid10 ssd and SAS - NL's raid6. or are they built raids more as how 3par manages the CPG, creating mini-raids out of pieces of different discs?

  HP also has a wealth of easily accessible information about the paintings of 3par, I can't find anything for compellent :(

  http://h20566.www2.HP.com/portal/site/hpsc/public/PSI/manualsResults?sp4ts.Oid=5044394&AC.admitted=1388181376886.876444892.199480143

  OK - if really it's better done in person with someone who can demo and animate the allocations of blocks for you.

  BUT

  Compellent storage virtualized Tier - so, for each level of storage (which are defined by the disks of like I/O features such as speed control, flash type, etc.) all of the devices used to treat each IO.  HDS let each LUN to have several RAID algorithms in use at the same time. The ideal on this would be to have write operations occur in RAID 10 (fastest to write) and read operations to get the RAID5/6 (more effective space). RAID types reserve NOT a specific amount of physical disks - but rather function as a method of organizing for blocks of different LUNS across all the disks in a pane.

  Data Progression (Tiering) extends this feature to allow volumes exist with blocks into several levels based on age and the activity of each block. The LUN is presented as there is no difference for servers based on the location of the data and the blocks can be shifted dynamically as criteria. Recently given Progression has been expanded to allow the mixing of writing optimized and read optimized Flash SSD levels that revolutionize the market tables around without resorting to the use of SSDS as a front-end cache. Data is transferred more quickly, then a normal progression among the readers of optimized for reading writing optimized SSD to keep performance levels greater than 100 K IOPS / s while valuing the cost differences inherent in MLC drives (they are cheaper).

  http://www.Dell.com/learn/us/en/qto08/shared-content~data-sheets~en/documents~Dell-Compellent-software-suite.PDF

  One thing that really sets storage Compellent, is that the license is NOT related to the material. If a controller SC8000 can be replaced without purchase of multiple licenses or invalidate the original license. The license is perpetual and abstract the hardware that is used. So if I have a HDS I bought 5 years through a license for all the features for 16 disks I can apply this license to SC8000 controllers and SSDS (up to 16 spare non-pieces) and start using it.

  I think that if you use this site as your jumping off point - http://www.dell.com/learn/us/en/04/dcm/flash-storage you will find a lot of information just look at software storage Center rather than the specific hardware building blocks.

• The RAID controller compatibility

  Hi I just bought a new system and intend to use a RAID controller to configure a RAID array.

  The controller that is available is ""Intel RS25DB080"." I've never done this before so to start I wanted to ask you how can I check if it is compatible with my x 99 Deluxe motherboard?

  (he did not appear on the page that mentions the compatible products: Specifications of RS25DB080 for the Intel® RAID controller)

  Secondly, it would be possible to set up several RAID arrays using a single controller? that is, if I want to do a table composed of 3 disks and another 3 more I have to buy an extra card?

  Also if this controller is not compatible what other controller I'll buy? And what exactly what I look for in a RAID controller if I'm looking for one? (Currently I just check to see if it's a SATA or SAS controller)

  So, basically, do not know much about the RAID needs your help, thank you!

  Generally if you are serious about RAID for the only real video editing solution is arec, because you will need 3 RAID or Raid 6.  There are a number of X 99 different systems with the Areca RAID card installed I am owner of a BOW 1883i, which works very well in my x 99 motherboard, I have it configured with 7 discs each RAID 3 Samsung 840 Pro I probably will be getting rid of it I can get nearly the same results with a SSD of unique m2.  In my old age, I went basic/simple rather than exotic.

• Problems with the raid PERC S100 controller

  Hi all

  Recently, I bought a Dell R410 server with a raid PERC S100 without search controller before that. Problem is that I need to install a Linux OS and as it seems, this controller does not support linux.

  Because I was in a hurry, but they can also help in my situation, I bought the following cables:

  

  Apparently, I don't have a slot for connecting the cables... :)

  So my question is. Is there a solution to this problem? I tried to disable/remove the S100, but the MB does not detect hard drives. The server is with a bottom of basket.

  I actually found this thread in the forums-> http://en.community.dell.com/support-forums/servers/f/906/t/19472600.aspx, but unfortunately, I couldn't find the cables posted in the thread.

  Here is an actual photo of the machine:

  

  This server will support the SAS 6/ir, Perc 6 / i and controllers Perc H700. you will need to run linux on this server. The SAS 6/ir is limited to RAID 0 and 1; Perc run several RAID levels.

  If you get the SAS 6/ir or Perc 6 / i, you will need this cable:

  69V70 cable, 6i/SAS PERC 6i/5 at the bottom of basket of 6G

  For the H700, need you also this cable:

  N262J cable HARD drive; Map PERC H200/H700 b/p (Mini SAS to Mini SAS

  Concerning

• ASM vs San raid which to adopt

  Hello
  
  We have a system using ASM and San drive. San has a hardware raid, while the DSO has also do the same thing. Because according to my understand its function. Strip and any mirror. Do you think we have several raids in the system and we should disable hardware raid. Also I thing hardware raid is raid 4. Please correct me if I'm wrong
  
  concerning
  Nick

  RAID 1 + 0 or 0 + 1 is an implementation of striping and mirroring.
  RAID 5 is an implementation of aggregation by band with parity (where the parity strip is scattered between the discs, with RAID 4 parity stripes are on specific disks)
  See: http://www.baarf.com

  Please, mind, than the baarf Web site focuses on the decrease in performance (writing) of the use of RAID levels with parity. Even with the SAN, it is still quite true. (it is inherent on the way to parity is implemented)

  ASM normal redundancy is essentially a mirror application. The implementation of normal redundancy ASM subtly different way works mirrored RAID.

  This means having RAID 1 + 0 / 1 + 0 on SAN level is a way of mirroring, and having normal redundancy ASM is another way to have copies mirror blocks, which would mean that having RAID mirroring on the normal SAN and level of redundancy at the ASM level means you have 4 copies on the same storage box.

  So I would say that there is little benefit to have normal redundancy ASM on top a stripeset mirrored.

  As an adviser what to do: most ASM implementations use external redundancy, which means the redundancy of the San is used. I think it makes sense.
  With the help of normal redundancy makes sense when you use local drives (non-RAID), or when having multiple SAN.

• Limitation of UCSM integrated server C-Series

  I crossed some line of integrated UCSM Server C-Series

  http://www.Cisco.com/en/us/docs/unified_computing/UCS/c/SW/RAID/configuration/guide/Cisco_UCSM.html#wp1068282

  Maximum Volume a RAID and a single RAID Controller in the built-in Rack-mounted servers

  A rack mount server that has been integrated with Cisco UCS Manager can have a maximum of a RAID volume, regardless of how many disks are present on the server.

  All local hard disks in an integrated rack server must be connected to a single RAID controller. Integration with Cisco UCS Manager does not support the connection of local hard drives rack-mounted several RAID controllers on a single server. We therefore recommend that you ask a single configuration of RAID controller when you order rack mounted servers to be integrated with Cisco UCS Manager.

  In addition, do not use third-party tools to create a multiple RAID LUN on rack mounted servers. Cisco UCS Manager does not support this configuration.

  Wdey,

  The following comment in the documentation is inherited an error early in UCSM when local disk policy was not well understood.

  "In addition, do not use third-party tools to create a multiple RAID LUN on rack mounted servers. Cisco UCS Manager does not support this configuration"

  It was added so users would not use WebBIOS and then set a policy on the server that would crush like manual mentions.   We that will remove the documentation.

  The method supported by Cisco for the creation of multiple RAID volumes on a built-in server of series C are like manual Describes.  Use WebBIOS to create disk volumes and create a strategy of local drive of the 'any configuration.  It is also recommended that all strategies drive on your system is configured with protect the configuration to prevent overwriting an existing configuration, unless the political action you want to perform.

  Steve McQuerry

  UCS - Technical Marketing

• SQL and Exchange on the same material - need help with storage!

  Hi all!

  I am definitely a newbie in the world of virtualization.  A colleague and I have spent hours scouring the Internet and reading white papers for answers and have been unable to find exactly what it takes.  So far, there has not been a lack of trying.  We are a very small company with 75 employees.  We have an Exchange Server and implement our first SQL Server and we want to virtualize both on hardware nine using the ESX 4.0 Essentials.  I don't know, probably not the best two apps to virtualize on a single piece of hardware, but we have about 12 people, using SQL server... at least for the foreseeable

  the future.  I'll try to make this simple in my questioning:

  (1.) what is the difference between a logical unit number, a PIN, a VMFS, a VM data store and how they relate to each other?  I know what means LUNS and VMFS and I know what a brooch is a hard drive, so no links to wikis on the definition of a LUN ;-).  Been there, done that.   These terms seem to be used very freely in all directions and it's confusing about how they relate to storage.

  (2.) we have 6 146 GB SAS drives for storage on my HP ML350G6, the config that I want to work with is the following:

  2 hard drives - RAID 1 - held Exchange OS, T-logs Exchange and SQL OS (possibly install VCenter Server on that as well in another virtual machine)
  4 hard drives - RAID 10 - will hold the Exchange and SQL DB database.

  Y at - it faults with this configuration?

  (3.) if #2 above is OK, how you (and you should in our small store scenario) cover several RAID (spindle?) arrays with a virtual machine.  In other words, how to you create a virtual Exchange computer and upgrade the operating system on a RAID, newspapers on the same value but different partition RAID and the Pb on a completely different RAID set?

  (4.) if #2 not OK, I should just create 2 simple VM Exchange, a for SQL and worry not separate newspapers/operating systems of databases on RAID berries?

  I thank you very much in advance!  These forums are wonderful!

  Confused in Cincinnati,

  -Bryan

  Thanks for your reply Wes!   I have a few questions for follow-up

  make sure that I understand.

  Therefore a unit logic simply number a "partition" carved on a storage space (if that)

  storage is a single hard drive or a RAID array)?   And of course each hard

  drive is a PIN, but when you take 3 discs and their RAID together.

  they, in fact, become a PIN, especially when it comes

  on reduncanc and no performance?

  VMware "" still does not see the storage, particularly the way in which you install. ESX will simply see a drive, format it with VMFS, so you can create files of your virtual hard disk on top of that (.vmdks) and your virtual configuration (.vmx)

  I'm not sure I understand the concern with VCenter Server?   It can not

  be on the same box as ESX running a VM couple?   My opinion is that

  VCenter Server would be just an another VM on this box?   Maybe I need to

  Read more on the VCenter Server!

  vCenter may work in esx, but it is preferable to have an another box, also don't forget that vcenter has its own SQL database and needs 3 + GB of ram. In your configuration, you can do better leaving vcenter out of the picture completely.

  Thanks again!

  -Bryan

  More like this:

  RAID 1 array: LUN1-> Datastore1-> ExchangeOS.vmdk, ExchangeLogs.vmdk, SQLOS.vmdk RAID 10 Array: LUN2-> Datastore2-> ExchangeDB.vmdk, SQLDB.vmdk

• The local disk on the ESX Server 4.0 workspace

  I have a Dell 2950 GEN III 2 x Quad Core 2 GHz 16 GB Ram, and 6 x 2 TB SATA hard drives and cards 6 network

  I am only able to create a unique database small.  I have 6 drives in a hardware Configuration RAID 5.   How can I use 100% of apx 10 TB of disk space local invited several flavors of Linux and Windows.   I understand that I will probably have several data stores.

  Do not hesitate to ask questions, I look forward to working this to enjoy this local storage

  This issue I'm having is install ESX is 1 TB, why does the rest of the free space not appear because it is all a local drive in the great hardware but I do not have enough space to do a data store apx RAID 5 via the SATA controller.

  Thank you

  If it's ESX ESXi4.0

  try it with several RAID 1 volumes and create multiple VMFS datastore each with size of 2 TB

  concerning

  Maniac

• Hardware iSCSI SAN SMB solution to possible

  We are looking to the UltraStor RS8 IP-4 by improving technology. Please take a look.

  http://www.enhance-tech.com/products/UltraStor/RS8_ip-4.html

  My main question for this discussion is... Why are they not everywhere? Looks like the perfect with a price less $5 000 CAD. Four ports of 1 Gb, port management, redundant PS, Snapshots, RAID 5 or 6 with the Hot Spare disks. He uses ordinary hard drives and they can even make if necessary SAS disks. VMWare has certified the RS16 IP-4 for 3.5 too, so I guess the 8 Bay must work found in vSphere 4. Am I missing something?  They have even a "how-to" to connect the SAN to the ESX host. See the attached pdf.

  Sounds too good to be true. Someone of you know something I don't know? What I forget? These are not spoken anywhere.

  Otherwise, we will watch a MD3000i from Dell for double the price at least.

  FYI, we will initially have 2 ESX host servers connected to it to VM - HA to the license of vSphere 4 Essentials more.

  I welcome all your comments.

  My thoughts on this vs the MD3000i.

  1. 8 drives in 2U is not look too good when you can get (15) or (16) 3.5 "in 3U or drives 2.5" (24) in 2U.

  2. A single controller RAID vs dual is a point of failure.

  3. Does not say SAS anywhere in the spec, just the SATA, so it is questionable

  4. Said only 2 GigE ports on the card rather than 4 as in the description, it is questionable as well.

  5. Support contract?

  I'm willing to bet that the tightening of the bolts 3 & 4 are a few typos when they have updated the product, but the tightening of the bolts 1, 2 and 5 are breakers of deal for me.  I need to cram in as many axes by RU as I can, several RAID controllers means greater availability, and when it does not break, I need someone here immediately with spare parts (or if necessary complete replacement unit), not a guy randomly on the phone of the India.

  Jason Litka

  http://www.jasonlitka.com

• Trying to build this mid/high range PC for prepro & AE

  Advance thanks for your comments and your tips for my next machine.

  I have spend many hours on this forum and came with this Conference.

  Fractal design Define R3 (case bigtower)

  OCZ zx 850 gold

  ASUS P8Z68 deluxe mobo (easy overclocking, I think that... is possible 4,.8 to 5.2 easy?)

  CPU Intel 2600 k

  Rev of Scythe mugen 2. B CPU cooler

  LG BH10ls30 BR burner

  4 x Hitachi Dekstar7K3000, 3 TB hd (images) in raid 10, less demanding on the cpu compared to raid 3 or 5? or not with the piece Z68?

  2 x WD Velocirapter 150 GB (1 for the OS, 1 for the pagefile, scratch)

  (exit/back up via Esata for external drives)

  MS professional 7

  Kingston 16 GB DDr3 1600 Hz kit

  ZOTAC Geforce GTX 480 1500 MB of memory

  What would you change?

  Thanks again for your time and your knowledge!

  Luc van Vliet

  Luke,

  What can you have several RAIDs using motherboard, Intel ports SATA ports, I can say Yes. The Board you get has 6 ports (4 Sata II, 2 Sata III) Intel and I'm sure you could even make a single drive RAID 6 baptized until going to Sata II speed.

  I thought:

  -C: VR 150 GB OS/programs - on one of the two ports Sata III of Marvel

  -P: 4x3TB RAID 10 array (projects, media, data, etc) - on 4 Intel Sata II ports

  -S: 2x1To RAID 0 (zero, rendering, etc) - 2 Intel Sata III ports (does not have Sata III speed, it just seems like a way to put in place everthing)

  See you soon,.

  Jim

• HP Z840 workstation: for Z840 liquid cooling?

  I currently have a Z800 WS dual Xeon CPU and coolant.  I am scheduled for a new WS and watched the Z840 but I don't see an option of liquid cooling. I run the structural analysis software that allows me to analyze a load case by heart so more cores means a faster flow.

  I was told that the faster clock speed and multiple real cores will be the best, so what will work better CPU and RAM?

  Another question, is that HP has a Mobo that supports several consumers i7 cpu popular with a lot of ram? I don't know if the Z840 WS and processor Xeon is the best bet.  Are there other options?

  There is no option of liquid cooling for z840 systems. With current processors, we do not believe that the system requires liquid colling option. HP get a kit rather than high-performance cooling liquid cooling in air. I don't know when, but not a long time.

  The following quicspecs list the supported processors and the memory modules. This system supports INTEL® XEON® E5 and DDR4 - 2133 ECC Registered / RAM load reduced (LR).

  http://WWW8.HP.com/h20195/v2/GetPDF.aspx/c04400043.PDF

  For performance, it is not a simple question. It depends on your application such as processor, memory intensive, drive-training or other resource intensive applications.

• Z840 HP: HP Z840 CPU

  Pending an answer to my question of cooling liquid, I will ask another question.

  The line processor Xeon E5 is the fastest CPU listed for the Z840. It looks like a Xeon E7-8891 is faster and I wonder if it might be available in the Z840 in the near future? Any plans for a Xeon E7 for the Z840?

  GTWatson77459 wrote:

  "With my situation in mind, what would be the best compromise of clock speed and the number of cores of the Z840?

  It's an interesting question.  The short answer is that it is application specific, but it is difficult to determine how to balance carrots vs. frequency for each application.  I don't really know the answer, but there are several variables to consider.

  The Z840 supports 3.5 GHz 4 core 2.3 GHz 18 of the CPU core.  Intel numbers according to their idea of relative performance processors, i.e., E5-2699 is the 'up' and the E5-2603 to 1.6 GHz and 6 hearts is the lowest.  HP QuickSpecs Z840 are here.   This gives a rough idea of the relative performance.

  Some points of reference, such as those on cpubenchmark.net, have higher ratings for processors with the majority of souls.  It's a way to rate CPU.  But we know as high-frequency players less lower cores.  The Z840 offers the processor 3.1 GHz 10 E5-2687W, which is a CPU high power (160W) Workstation only.  This could be the sweet spot, but don't take my word for it that it will be the best for your application.

  A little off-topic technical discussion:

  Frequency rated core from Intel is the guaranteed minimum of frequency.  All processors have the Turbo Boost technology, which increases the base frequencies by steps of 100 MHz.  CPUs are given numbers for turbo steps maximum 'all basic' and 'carrot' turbo maximum steps.  For example, the E5-2699 is only 2.3 GHz, but can stimulate all the hearts in 5 steps and a carrot in 13 stages, under certain conditions.

  Intel found that the maximum power of the CPU is limited by the temperature of die.  Too hot, and the industry may be damaged (in fact it is thermally first throttle) so Intel limited the power being dissipated.  (Each processor has a rated output of Groove).  Die temperature corresponds to the use of the power of the CPU.  Use your analogy (of cleaning units = pressure x Volume) pressure washer, similarly, CPU power dissipation is a function of: [frequency of the processor (higher freq = more power used), the number of active nuclei and the load on each core].  The idea is to run the CPU at a high level of power for maximum performance, so the frequency of the hearts will vary depending on the load.

  If the system uses only 1 core, temperature of die will get too hot, if there is an another carrots 15 just sitting there.  In this case, Intel will increase the active base of several steps of 100 MHz frequency.  10 stages = 1 GHz, so what rated 2.3 GHz CPU could run to 3.3 GHz for a short period.  If all hearts work 100%, the hottest of, so the frequency of operation may be limited to the nominal value.  The frequency may change the order of milliseconds, if I remember correctly.

  Another mechanical analogy, it is that of a car with a redline 6000.  Maximum power (= performance) is then carried out to redline, even, the CPU performance maximum occurs at the maximum temperature limit.  If the CPU can run at up to x Watts, why not increase the frequency when he's not stressed?

  It is a simplification, but tries to explain the operation of the CPU.  Sometimes people post in the forums that their CPU temp is high, and they are worried about it.  As long as the temperature is lower than the maximum junction temperature Tj max, it's OK.  (Running coretemp program displays Tj max and the base time, FYI).  If the system is working properly, with the correct heater and the circulation of air, it runs the warmer is a sign to get the best performance from the system.  Running a car to redline makes it go faster, but it heats up.  ;-)

  BTW, in the same way, if the CPU usage is very low, the frequency of the processor will drop a lot of steps, sometimes a GHz or more.  This is done to save overall system power when the processors don't make much.  Run an active program of CPU frequency will show bounce everywhere.

  The UC E5-2687W is the only supported processor 160W.  Intel suggests that it's the highest "balance" CPU because it has a relatively high frequency of 3.1 GHz to 10 cores and has more thermal height.  But even once, do not take my word for it that it is better for your application.

  HP found that some programs do not recognize more than 64 hearts, so 2 x 18 E5-2699 processors cores with hyperthreading, will use only 64 out of 88 cores available (with active Hyperthreading).  Just another complication.  FOR INFO.

  Sorry for the long post, but I hope that some of it is useful.

• BIOS Configuration Utility (BCU) for HP Z840 workstation.

  Took a look in the download section for the workstation Z840 but could not find. Someone has a link to this software need a 64-bit version and must be compatible for the new workstation Z840.

  P.S. This thread was moved from workstations - z dad series, series xw and legacy x and w series
  HP PC Client management. -Hp moderator of the Forum

  Never mind found a link to get the latest version.

  http://FTP.HP.com/pub/caps-SoftPaq/CMIT/HP_BCU.html