EPM and OBI SSD type HDD optimization

How are we to configure our EPM and OBI suites if we are to optimize SSD type hard drives. When it comes to SSD drives, the rule is the fewer writes the better. One can probably prolong the lifespan by 25% if one make necessary optimizations.
Server specific optimization:
*Drive indexing disabled (useless for SSD:s anyway, because access times are so low).
*Prefetch disabled.
*Superfetch disabled
*Defrag disabled.
Virtual Machine host
*Store VM-files on a mechanic HDD while the SSD is used solely for the VM-machine's (virtual) HDD-file(s).
EPM
*Essbase?
OBI
*x?
I'm about to order two 60GB generation 3 SSD drives with read speed of 550MB/s and write speed of 475MB/s and 50K IOPS. I'm really curious on the improvement in system performance this will entail.
Edited by: Markus on Aug 10, 2011 10:55 AM

Hi Fraser82,
Thank you for visiting the HP Support Forums and Welcome. I have read your thread on your HP Notebook and not seeing the HDD. Here is a link to update the BIOS.
I would be happy to assist if needed. How Do I Find My Model Number or Product Number?
Please respond with which Operating System you are running:
Which Windows Operating System am I running?
Please let me know.
Thanks.
Please click “Accept as Solution ” if you feel my post solved your issue, it will help others find the solution.
Click the “Kudos, Thumbs Up" on the bottom to say “Thanks” for helping!

Similar Messages

  • MSI GT60 Windows 8 - SSD Raid is detected as HDD (Optimize)

    So I just got the GT60 and I have my OS installed on my Super Raid 128GB setup.  The problem I have is that when I went to the "optimize" option of Windows 8, I saw that it detected my SSD Super Raid Setup as HDD and not SSD.  I thought analyzing the drive might fix it, but it didn't.  Now I'm afraid to optimize it for fear that it will defrag it instead of TRIM.
    Edit: Somehow after 2 days of not checking on it, it somehow fixed itself... I don't know how it did it, but I didn't do anything to it other than some Windows updates and app crash fixes.  I guess that fixed it...
    Well as a side question, should I optimize the other 2 drives that I see?
    BIOS_RVY
    WinRe Tools
    Should I run optimization on them?

    Original devices? What devices?
    My stripe size is 128kb.
    I have used other benchmarks. Unfortunately mark the same results.
    I have also looked at the firmware of the SSD are updated. I tried one without ssd raid, and I get 200mb result of reading and writing 90MB.
    Something is wrong, why results so low?
    PS: Seen this way, the raid is making me x2 speed, but why the low result?

  • After clean install/SSD upgrade, how to keep prefs and wipe OS from HDD?

    I'm about to install an SSD in the boot bay and move my HDD to the optical bay.  I'm awaiting the arrival of the disk doubler before I get on with it. In the meantime have a couple questions:
    Firstly, after I've done a clean install of Mountain Lion and loaded up all my apps onto the SSD, I'd like to free up memory and remove the OS and apps from the HDD.  After googling, I've found lots on installing SSDs and creating a USB boot drive with ML, but can't find anything that addresses potential issues of simply deleting the OS from the old (HDD) drive.  I'd appreciate any advice on dos and don'ts!
    Secondly, is there any way that before I delete these, I can import all my existing prefs/settings for the OS and existing apps, so I don't have to reset all of this?  Again, despite looking around for tips, I've come up empty.
    Thanks in advance for any help.
    GG

    Anyone?

  • Partitioning 500GB-HDD and 80GB-SSD for Arch

    Hello guys,
    I have a laptop with a 500GB Harddisk and a 80 GB SSD installed, and I would like to get the most efficient partition layout.
    Note: Many times I have seen people suggesting putting the swap or the /var/tmp on the SSD, that is definitely something i do not want. SSD Disks have a limited number of write operations before they die, and therefore putting the Swap to the SSD is ridiculous.
    I have come up with the following partitions scheme, and i would like to discuss, and refine that with you:
    --- To make things easier I called the HDD hddX and the SSD ssdX. ---
    /dev/ssd1        /boot            ext2        noauto,noatime    1 2
    /dev/ssd2        /            ext4        noatime        0 1
    /dev/hdd1        /var/tmp        ext4        noatime        0 0       # on hdd for compilation
    /dev/hdd2        /home            ext4        noatime        0 0
    tempfs            /tmp            tmpfs defaults,nosuid,size=1024M,mode=1777    0 0
    16 GIG Swap at beginning of HDD!
    Is there any refinement you see? Or anything else in /var that must go on the HDD?
    Best Wishes,
    Daniel
    Last edited by alphamaennchen (2011-05-31 05:37:52)

    I have been using ssd laptop for quite some time now, and I have to definitely agree with graysky about swap. You probably don't need swap. It's an echo of the past, when PC's really were low on ram... Today, unless you use some specialized memory-hungry software you can do without swap even with 2GBs of ram. Check your system monitor, it's likely that you have a big memory redundancy all the time. And 16GB swap is ridiculous - you would go mad with software working on swap, even 2GB, believe me. And running all that you have in your system probably wouldn't use all your ram. It's funny how people still think that they need swap today. Use it only if you use hibernate.
    I would put /boot and / on ssd. I don't see any reason to put /var on ssd - surely ssd is fast, but hdd isn't that slow - I doubt that var would use more than 50MB/s, which every modern HDD should exceed. And depending what you mainly use /home for, it also can be put on ssd.(/home is hmmm... "unpredictable" - for example WINE can mess it up).
    And remember to add noatime and discard mount options to ssd partitions. There are few other tweaks for ssd too. For example, if you put /home on ssd, you can move your web browser's cache(I moved it to RAM).
    Last edited by bart_vv (2011-08-02 20:51:27)

  • Linux on SSD and Windows XP on HDD

    This is what I am planning to do: Install Linux on SSD and Windows XP on HDD. It is a constant problem that the SSD fills up too fast. And when downloading from the internet it is not always given as an option to install on other drives than C:, for instance Adobe. So I will install Linux on SSD and maybe use some of the 160Gb hardiskspace as a logic partition for Linux. The rest of the harddisk will be C: and have Windows XP on it. The idea is to have a quickstarting Linux-pc with Firefox for internetuse only, and a regular Windows-pc. 
    But before I do this, is it anyone out there who has tried this? Anyone with warnings, cautions or notes? Recommendations? Or is it just not possible?
    Polaris

    why not install xp on the hdd and use lenovo quickstart for fast surfing using firefox? this will save effort and disk space..
    AMD Phenom II X6 1055T / 2xATI HD5770 CF / 8GB DDR3 / 4x500GB RAID-5 / Win 7 Ultimate x64
    Lenovo S10e Black / 2GB DDR2 / 160GB HDD / 6-cell battery / Win 7 Ultimate x86
    Motorola Xoom

  • Satellite U840W PSU5XA - Win 8.1 installation and effective SSD usage

    Hello.
    Im hoping for a solution to two issues.
    The second I suspect is caused by the first.
    Satellite U840W PSU5XA-001006
    Started to receive BSOD with various faults.
    Most persistant was IRQ LESS or not EQUAL- iaStorAV.sys
    Many hours of searching and I cannot locate a solution nor post that actually has the "AV" extension on on iaStore.sys.....
    So, i factory resetted, and various other issues (BSOD's) persisted.
    So, I Clean installed Windows 8.1 straight over everything and deleted all previous files.
    I noticed the system was not running any factory TOSHIBA Branded utilities etc. and seemed to be a lot slower. I suspected that this was caused due to the 32GB SSD used for Cache had now seemingly been bi-passed and everything was now strictly running off the 500GB HDD. I dont recall BSOD's but the system was significantly slower on boot and performance.
    Sooooo, (SIGH), I again reset back to factory settings and this gave me back TOSHIBA branded utilities such as service station and the like.
    HOWEVER, The BSOD's remained and once upgrading to windows 8.1 KEEPING files and settings, the iaStorAV.sys error has returned (specifically when closing lid into sleep mode) forcing CRITICAL error and forcing system into reboot instead of awakening from sleep.
    Note: Windows is a legit MS copy so that isnt the issue either. Even stranger is that the 8.1 instal had been running at least 8 months no issue prior to sudden BSOD Page file and iaStorAV errors. Readings and posts have me believing MS Updates to bee the culprit here for the beginning of these issues.
    I read a similar post regarding enabling RAID CACHE PRIOR to Windows 8.1 install. My intuition tells me that these two may just be related. i.e. If windows has not installed RAID CACHE properly, then CACHE isnt being strored correctly and causing BSOD's. Im a newbie at this stuff, so be patient.....
    What I really want to know is, How can I now Factory Reset Windows 8, making positive the 32GB is being used as CACHE and assisting fastboot etc, then safely upgrade to Windows 8.1?
    Any assistance would be greatly appreciated.
    THINGS ALSO ATTEMPTED: IRST and RAID re-Install, Memory Diagnostic (Passed), BIOS Upgrade to 6.70, Rapid Storage Drivers re-installed
    Thanks ahead of time,
    Mr Frustrated............

    Ok, well, its been a while and no-one had really come up with a solid solution. But thank you JakieStewart for the RST information.
    Note: Updating RST drivers alone, failed....
    However, I have been able to work around the issue in a way:
    I have lost the ability to use the SSD as a cache, which really annoys me, as all the reboots, and suggestions i have read, re-installs etc proved flawed in some way or another.
    So, I have instead chosen to clean install windows 8.1 then followed a post relating to Diskpart and setting SSD hibernation partition.
    A link to consider would be http://download.intel.com/support/motherboards/desktop/sb/rapid_start_technology_user_guide.pdf
    However i followed a somewhat abbreviated version.
    Conclusion:
    1. Now I can close the lid on my U840W (INTO SLEEP or HIBERNATION Mode and not get a "System Thread Exception iaStorAV.sys error because I have 1. Reluctantly Clean Installed Windows 8.1 over factory image and 2. now created a >Hibernation Partition< on the SSD
    Even though SSD Cache may no longer be available, creating this Hibernation Partition assists with Rapid START Technology and the associated iaStorAV.sys type errors
    You can find your Toshiba Rapid Store Technology Drivers for your model here:
    https://downloadcenter.intel.com/search?keyword=Intel%C2%AE+Rapid+Storage+Technolog y+%28Intel%C2%AE+RST%29
    2. I no longer have the option to use the SSD as a cache, but overall my system runs a lot better now
    3. The SSD is not visible in explorer (which it never was under the original settings anyway)
    3. Downside: I can no longer boot or reset back to original Toshiba factory settings and can only recover from the new recovery disk (USB) i have made.
    I have now also downloaded and stored on a USB all the original TOSHIBA U840W Drivers from http://www.mytoshiba.com.au/support/computers/ultrabook/u840w/psu5xa-001006/download?os=30
    If like me, you clean install windows 8.1 I highly suggest getting your drivers downloaded locally so that you don't have to keep going back to Toshiba or Intel unless updates are required.
    4. Note: Make note to set the BIOS Hibernation mode to "Immediately" if you choose to go this route as per the instructions on the Intel Link
    5. Change "Toshiba Power Options" to make sure that closing Lid send laptop into Sleep or Hibernation (Both work) Do this by searching for "power options' in windows charm search and >select power plan >Change power plan settings >Change Advanced Power Settings >Power Button and Lid >selecting either Sleep or Hibernation (both work as i said earlier)
    ON THE PRO SIDE: My HDD boot from total Shutdown State time is "about" 50 seconds but more significantly, closing the lid into sleep keeps my work available again, and takes 9.5 seconds to boot from sleep into usable actions!!!!
    So despite losing SSD Cache, I now know a lot more about my computer, and the sleep function and Hibernation work supremely well.
    I hope someone finds this thread useful...
    Shane
    Laslty, it is sooooo strange how this all happened overnight with the iaStora.sys; iaStorAV.sys iaStora.sys errors...and how restoring back to previous states didn't work, and how it seemingly corrupted even the Flash Stored Cache settings and Operating System! At least I now understand my computer better and luckily this didn't occur during previous Diploma studies. Phew.....
    Message was edited by: shaner1

  • SSD with HDD

    Hi
    I have a late 2011 Macbook Pro
    I am planning to buy a 240gb SSD for boot and application, and use the current HDD for file storage
    The HDD will be in the optical drive bay with a caddy, and the SSD will be where the HDD was before.
    I was wondering how EXACTLY I would go about doing this, right from the moment I put the hard drives in.
    Could you please include step-by-step instructions

    Install the SSD in place of the HDD. Use Internet Recovery to set up the SSD:
    Install OS X Using Internet Recovery
    Be sure you backup your files to an external drive or second internal drive because the following procedure will remove everything from the hard drive.
    Boot to the Internet Recovery HD:
    Restart the computer and after the chime press and hold down the COMMAND-OPTION- R keys until a globe appears on the screen. Wait patiently - 15-20 minutes - until the Recovery main menu appears.
    Partition and Format the hard drive:
    Select Disk Utility from the main menu and click on the Continue button.
    After DU loads select your newly installed SSD (this is the entry with the mfgr.'s ID and size) from the left side list. Click on the Partition tab in the DU main window.
    Under the Volume Scheme heading set the number of partitions from the drop down menu to one. Click on the Options button, set the partition scheme to GUID then click on the OK button. Set the format type to Mac OS Extended (Journaled.) Click on the Partition button and wait until the process has completed. Quit DU and return to the main menu.
    Reinstall OS X: Select Reinstall OS X and click on the Install button. Be sure to select the correct drive to use if you have more than one.
    Note: You will need an active Internet connection. I suggest using Ethernet if possible because it is three times faster than wireless.
    This should restore the version of OS X originally pre-installed on the computer.
    Now, you should be able to download Mavericks from your Purchases page in the App Store. Once done with this you use Startup Disk to set the SSD as the default startup drive. Once this is done, then you can install the HDD.
    Now, with the HDD installed you can redirect OS X to use the Home folder on your HDD:
    How to use an SSD with your HDD
    If you are going to use an SSD as a boot drive together with your existing HDD as the "data" drive, here's what you can do.
    After installing the SSD you will need to partition and format the SSD using Disk Utility. Then, install OS X on the SSD. After OS X has been installed boot from the SSD. Use Startup Disk preferences to set the SSD as the startup volume.
    Open Users & Groups preferences. Click on the lock icon and authenticate. CTRL- or RIGHT-click on your user account listing in the sidebar and select Advanced Options from the context menu. You will see a field labeled "Home dir:" At the right end you will see a Change button. Click on it. In the file dialog locate the Home folder now located on the HDD (HDD/Users/account_name/.) Select the folder, click on Open button. Restart the computer as directed.
    When the computer boots up it will now be using the Home folder located on the HDD.
    Another more technical method involving the Terminal and aliases is discussed in depth here: Using OS X with an SSD plus HDD setup - Matt Gemmell. This is my preferred approach because I can select which of the Home's folders I want on the HDD and which I don't want. For example, I like to keep the Documents and Library folders on the SSD because I access their content frequently.
    Be sure you retain the fully bootable system on your HDD in case you ever need it.
    I would not try installing and using applications from the HDD nor making any other changes. There is little to be gained by doing more than is described above.

  • Lenovo G560 - Issue with won't boot and black screen after HDD upgrade

    Lenovo G560 - Issue with won't boot and black screen after HDD upgrade.
     What happen: My laptop was working fine, no blue screen issue, no funny business at all. I bought a new SSD Intel 120 GB and thought it would be a good idea to replace the HDD. I shutdown and disconnected the power adapter, waited a couple of minutes and removed the battery. I opened up the back case and replaced the HDD. And put all the screws back and put the battery back in.
    Problem: The very first time I turned the power on, nothing happens besides a black screen. I pressed the dvd drive it works and opens up and closes. I waited for about 30 mins and still has blac screen. When I mean black screen, no bios menu, no logo, just a black screen with the fan sound on.
    I have tried these:
    1. Unplugged everything - battery, adapter, and pressed the power on button for about 60 seconds, nothing, the laptop turns on with the LED display on for both on and battery LED's. But nothing but black screen, no sound of windows loading just the fan and black screen.
    2. I tried putting back my old 2.5" and nothing but black screen.
    Thoughts and suggestions?
    Solved!
    Go to Solution.

    Hi Autoexit173,
    Welcome to Lenovo Community!
     As per the query we understood that you are facing issue with system not booting in your Lenovo G560 laptop.
    As you have mentioned that the system not booting, please try to remove the RAM and  turn on the system and check if you can hear any beep sound. Also try to clean the RAM slots and check for the issue.
    Click here for the steps to remove the RAM and refer page number 40.
    Hope this helps. Do post back if issue persists!
    Best regards,       
    Ashwin.S
    Did someone help you today? Press the star on the left to thank them with a Kudo!
    If you find a post helpful and it answers your question, please mark it as an "Accepted Solution"! This will help the rest of the Community with similar issues identify the verified solution and benefit from it.
    Follow @LenovoForums on Twitter!

  • Re: How do you create and use "common" type classes?

    Hi,
    You have 2 potential solutions in your case :
    1- Sub-class TextNullable class of Framework and add your methods in the
    sub-class.
    This is the way Domain class work. Only Nullable classes are sub-classable.
    This is usefull for Data Dictionary.
    The code will be located in any partition that uses or references the supplier
    plan.
    2- Put your add on code on a specific class and instanciate it in your user
    classes (client or server).
    You could also use interface for a better conception if needed. The code will
    also be in any partition that uses or references the supplier plan where your
    add on class is located.
    If you don't want that code to be on each partition, you could use libraries :
    configure as library the utility plan where is your add-on class.
    You can find an example of the second case (using a QuickSort class,
    GenericArray add-on) with the "QuickSort & List" sample on my personal site
    http://perso.club-internet.fr/dnguyen/
    Hope this helps,
    Daniel Nguyen
    Freelance Forte Consultant
    http://perso.club-internet.fr/dnguyen/
    Robinson, Richard a &eacute;crit:
    I'm relatively new to forte and I'd like to know how can you handle utility
    type classes that you want to use through out your application? Ideally
    what I want is a static class with static methods.
    Let's say that I have a StringUtil class that has a bunch of methods for
    manipulating strings.
    My problem is that we have code that runs on the client and code that runs
    on the server. Both areas could use the StringUtil class, but from what I
    understand, I have to create StringUtil in a plan and then create a server
    object of type StringUtil. The server object will eventually get assigned
    to a partition. That's not good since I really want the server object to
    physically reside at the server end and at the client end. (Actually, I
    don't want a server object, I just want to invoke a static method of a
    static class).
    Any clues on how to solve this problem would be appreciated.
    Also, what is the url at Sage-it that has a summary of all emails that have
    been posted to [email protected]? Perhaps this question has been
    answered previously.
    Thanks in advance
    To unsubscribe, email '[email protected]' with
    'unsubscribe forte-users' as the body of the message.
    Searchable thread archive <URL:http://pinehurst.sageit.com/listarchive/>-
    To unsubscribe, email '[email protected]' with
    'unsubscribe forte-users' as the body of the message.
    Searchable thread archive <URL:http://pinehurst.sageit.com/listarchive/>

    Hi Richard,
    Your question about "utility classes" brings up a number of issues, all of
    which are important to long-term success with Forte.
    There is no such thing as a static method (method that is associated with a
    class but without an implicit object reference - this/self/me "pointer") in
    TOOL, nor is there such thing as a global method (method not associated
    with a class at all). This is in contrast to C++, which has both, and
    Java, which has static methods, but not global classes. Frequently, Forte
    developers will write code like this:
    result, num : double;
    // get initial value for num....
    tmpDoubleData : DoubleData = new;
    tmpDoubleData.DoubleValue = num;
    result = tmpDoubleData.Sqrt().DoubleValue;
    tmpDoubleData = NIL; // send a hint to the garbage collector
    in places where a C++ programmer would write:
    double result, num;
    // get initial value for num....
    result = Math::Sqrt(num);
    or a Java programmer would write:
    double result, num;
    // get initial value for num....
    result = Math.sqrt(num);
    The result of this is that you end up allocating an extra object now and
    then. In practice, this is not a big deal memory-wise. If you have a
    server that is getting a lot of hits, or if you are doing some intense
    processing, then you could pre-allocate and reuse the data object. Note
    that optimization has its own issues, so you should start by allocating
    only when you need the object.
    If you are looking for a StringUtil class, then you will want to use an
    instance of TextData or TextNullable. If you are looking to add methods,
    you could subclass from TextNullable, and add methods. Note that you will
    still have to instantiate an object and call methods on that object.
    The next issue you raise is where the object resides. As long as you do
    not have an anchored object, you will always have a copy of an object on a
    partition. If you do not pass the object in a call to another partition,
    the object never leaves. If you pass the object to another partition, then
    the other partition will have its own copy of the object. This means that
    the client and the server will have their own copies, which is the effect
    you are looking for.
    Some developers new to Forte will try to get around the lack of global
    methods in TOOL by creating a user-visible service object and then calling
    methods on it. If you have a general utility, like string handling, this
    is a bad idea, since a service object can reside only on a single
    partition.
    Summary:
    * You may find everything you want in TextData.
    * Unless you anchor the object, the instance will reside where you
    intuitively expect it.
    * To patch over the lack of static methods in TOOL, simply allocate an
    instance when required.
    Feel free to email me if you have more questions on this.
    At the bottom of each message that goes through the mailing list server,
    the address for the list archive is printed:
    http://pinehurst.sageit.com/listarchive/.
    Good Luck,
    CSB
    -----Original Message-----
    From: Robinson, Richard
    Sent: Tuesday, March 02, 1999 5:44 PM
    To: '[email protected]'
    Subject: How do you create and use "common" type classes?
    I'm relatively new to forte and I'd like to know how can you handle utility
    type classes that you want to use through out your application? Ideally
    what I want is a static class with static methods.
    Let's say that I have a StringUtil class that has a bunch of methods for
    manipulating strings.
    My problem is that we have code that runs on the client and code that runs
    on the server. Both areas could use the StringUtil class, but from what I
    understand, I have to create StringUtil in a plan and then create a server
    object of type StringUtil. The server object will eventually get assigned
    to a partition. That's not good since I really want the server object to
    physically reside at the server end and at the client end. (Actually, I
    don't want a server object, I just want to invoke a static method of a
    static class).
    Any clues on how to solve this problem would be appreciated.
    Also, what is the url at Sage-it that has a summary of all emails that have
    been posted to [email protected]? Perhaps this question has been
    answered previously.
    Thanks in advance

  • MacBook pro 13" /15" SSD type

    Hello,
    is here anybody in the forum who received their MBP´11 with an SSD? Maybe you can post the vendor and the type please? Maybe it´s an SandForce based SSD type ...
    Many thanks.
    Cheers
    tappi

    The label posted before http://i56.tinypic.com/10o4mty.jpg seems to have a bug ;-))
    Its mentioned THNSNC128GBSJ in the top, but if you look into the middle below the Logos you´ll find TSD-THNSNC256GBSJ(B)

  • When loading LR4.3 is it necessay to download to Drive c (SSD TYPE) rather than to drive D with 3T?

    When downloading LR4.3 is it necessary to download program to drive C with limited space (SSD type) rather than to drive D with 3T storage or will this not work properly (using a PC with Windows 8)?

    Let's straighten out terminology, shall we?!: It doesn't matter where you download it to, but it matters where you install it. The install will default to your system drive and while you can change the folder for the main part to point to your 3TB drive, some parts will always be installed on your SSD.
    Mylenium

  • SSD + slower HDD

    Despite the abundant discussions about good hard drive configuration for Lightroom (v 4), I am still struggling to understand what to choose when buying my new desktop PC without blowing my budget.
    Assume that processor, RAM, motherboard and all that are up to par, what should I go with in terms of HDD/SSD (under a somewhat limited budget)?
    I want a *quiet* system so using three roaring 10,000 rpm HDDs is no option. If anything I would like to use low power and low noise HDDs (5400 rpm), which could be combined with an SSD. Is this a stupid setup with the HDDs being a bottle neck? For example, is a 128 GB SSD (for Windows 7/64 OS, apps, LR cache) + 1 TB 5400 rpm HDD (for the image files) a good trade-off between performance and noise?
    Jalamaar

    I believe the impact of a green HDD is negligible.
    2012/8/1 Jalamaar <[email protected]>
       Re: SSD + slower HDD
    created by Jalamaar <http://forums.adobe.com/people/Jalamaar> in *Photoshop
    Lightroom* - View the full discussion<http://forums.adobe.com/message/4591485#4591485

  • [solved] How to install with old BIOS + 60GB SSD + 3TB HDD + fulldisk?

    I'm new to archlinux and this will be my first install. I use Debian and Kubuntu, but I'm no expert. I've read many archlinux docs (and they *are good*, congatulations), but I couldn't find specifics to what I want to do.
    My system is a desktop box (6GB RAM) that I use for video editing, compiling, browsing and many other unrelated things. What I plan to do is to organize things in the following way:
    SSD (60GB): first disk (for a fast boot up time) with only one partition, ext4
    / and everything, except for /home and /var
    HDD (3TB): second disk with only one partition, ext4
    Mounted as /something
    Symlinks (or bind mounts) to /var and /home
    My reason to use only one partition for / and one for /home and /var is to not have to worry about balancing partition sizes among system files, user files, web server files, etc.
    Could you please comment on the following:
    1) It seems that I can't use a a regular partition larger than 2.2TB, so I will have to use GPT. Do I need (or is it recommended) to partition both disks with GPT? Does it work with non EFI BIOSes? Can I use a regular GRUB2 install?
    2) What is the recommended way of installing arch using this set up? Do I install everything under the SSD and later move /var and /home to the HDD?
    3) What is the best way to have the complete final root file system: using symlinks, bind mounts, or any other technique?
    4) I deal with huge files and sometimes huge pictures and huge memory needs (Blender 3D, GIMP and Inkscape with huge files, radiosity rendering, etc). I've read that a swap partition is not needed nowadays. So I was thinking of creating just a swap file inside the HDD partition. Would it degrade too much the HDD performance or fragmentation? Could it be used for hibernation?
    5) Can I use the regular arch install procedure to accomplish all that or do i have to prepare anything before/after?
    6) Any other tips?
    Thanks
    Last edited by plenus (2012-04-21 02:49:47)

    plenus wrote:1) It seems that I can't use a a regular partition larger than 2.2TB, so I will have to use GPT. Do I need (or is it recommended) to partition both disks with GPT? Does it work with non EFI BIOSes? Can I use a regular GRUB2 install?
    According to wikipedia, linux has no problem booting from GPT on BIOS, and accordind to archwiki, GRUB2 provides the ability to boot from GPT in both BIOS and UEFI based systems.. I dont think that formating both disk with GPT is necessary, but I am not sure.
    plenus wrote:2) What is the recommended way of installing arch using this set up? Do I install everything under the SSD and later move /var and /home to the HDD?
    I would install with this approach.
    plenus wrote:3) What is the best way to have the complete final root file system: using symlinks, bind mounts, or any other technique?
    In this setup (/var and /home on one big partition), I would use symlinks.
    plenus wrote:4) I deal with huge files and sometimes huge pictures and huge memory needs (Blender 3D, GIMP and Inkscape with huge files, radiosity rendering, etc). I've read that a swap partition is not needed nowadays. So I was thinking of creating just a swap file inside the HDD partition. Would it degrade too much the HDD performance or fragmentation? Could it be used for hibernation?
    This says Google about swap file/partition performance. You could use hibernation. I allways go with swap partition, but nowadays it is just proforma, because I have enough memory to never use swap.
    plenus wrote:5) Can I use the regular arch install procedure to accomplish all that or do i have to prepare anything before/after?
    I think (but I'm not sure, though) parted is on installation media.

  • Reading while Writing on SSDs vs HDDs

    So, I've always heard that reading and writing to an HDD is not good, as it slows down the whole process (this is in video rendering for Premiere Pro).
    So, I know that SSDs are quite a bit different.  I'm assuming that reading and writing occurring on 1 SSD during rendering isn't going to be nearly as much of a problem as it would be happening on 1 HDD.  However, I'm unsure of this.  Am I correct in my assumption?  If not, what kind of performance degradation might occur, would it be noteworthy, and how comparable would it be to that of an HDD?

    Yes, in addition.....current high quality SATA III SSDs that have the best controllers in them  for video editing, ( like the Crucial M550, or, the more expensive Samsung 850pro), will not suffer from "fill degradation"....losing performance as the drives becomes more full. Also, the RELIABILITY ....especially on the Samsung 850 Pro....is WAY higher for the SSDs, than mechanical spinning drives.....allowing those quality SSDs to be considered for RAID 0 usage with a much lower risk.
    What you are getting at is the nature of the SATA interface itself......it is a "half-duplex" interface, which means it CANNOT write while READING...and vice-versa. It's like an old walkie-talkie radio....when you push the button to speak and transmit...you cannot RECEIVE anything ,until you let the button go. So, YES ,that "half-duplex" handicap is still there.....BUT, SSDs reduce its effect with their MUCH faster read and write speeds AND their drastically lower "latency", as Eric has said.
    There ARE places to connect storage devices, especially on the new Haswell E X 99 motherboards, which are "full duplex"...allowing reading and writing at the SAME TIME.  However, the storage devices to take advantage of this higher speed and full duplex connection currently are much more expensive than SATA devices. M.2, , SATA "Express", and PCI card storage devices can be used.....but, currently at high expense and limited capacity.
    With X99...one option is to take advantage of the huge amount of SATA !!! ports on the motherboard and just create RAID 0 with SSDs for storage..or, use cheaper HDDs, with a large "enterprise level" 7200rpm HDD to BACK UP the HDD RAID0 array !

  • 512 ssd vs 256 ssd 750 hdd 7200rpm?

    Hi I read that several people configure 256gb ssd + 750gb hdd 7200rpm, cause they use the  ssd for applications and the hhd for storage. But in general you
    will be working with both of them right? example whatch a movie. The program is on ssd but the movie is on hdd. So it will load faster on a plain 512 ssd or 2x256 ssd?

    Ralph your response is way off. The answer to OP's question mostly depends on write performance of the SSD, especially after the SSD has become full and the OS is starting to erase previously deleted (but not yet fully erased) blocks. That is when the SSD starts to slow down, and then the TRIM feature of SSD controller needs to be benchmarked. To answer the OP's question, one needs to run benchmarks of various use cases, especially against an aged SSD.
    I found this question myself googling for answers to same topic. Most benchmarks do not seem to be testing aged SSDs, but brand new out of the box SSDs that will usually smoke even 7200RPM drives. But what happens 3 months later after the SSD virgin blocks have been written to at least once and then it starts it slow erase cycles?
    Read performance of the SSD will always be faster than the 7200RPM drive, esp when comparing random access or fragmented file system accesses. So why does write performance matter if you aren't, for example, doing heavy write tasks like saving video files? Because the OS does tons of little random reads AND writes just sitting there... anything you do like open a new browser tab or start a Flash player will open up temp files for writing... so write performance becomes quite important if writes start to get really slow for any reason.
    edit: some anecdotal evidence that aging the SSD is not slowing down write for modern SSDs vs the bad experience I had with 1st gen SSDs.
    http://forums.anandtech.com/showthread.php?t=2218206

Maybe you are looking for