Slashdot Mirror


User: jabuzz

jabuzz's activity in the archive.

Stories
0
Comments
2,477
First seen
Last seen
Profile
(view on slashdot.org)

Comments · 2,477

  1. Re:How useful is this in realistic scenarios? on Open Source Deduplication For Linux With Opendedup · · Score: 1

    Your NetApp!!! Then you are already paying through the nose for storage.

    As for IOP's, *very* few machines ever get close to pushing the IOP's for one array, let alone the storage system.

    NetApp is about the only storage solution that provides dedup in the box at no extra cost. However per GB of storage it is one of the most expensive solutions out there.

  2. Re:How useful is this in realistic scenarios? on Open Source Deduplication For Linux With Opendedup · · Score: 1

    The point is though even if you save 20GB per OS instance, that only comes to 2TB over 100 virtual machines. You are talking of saving four RAID1 450GB 15k rpm SAS/FC arrays or eight disks. It really is just is not worth the additional complexity. At your 10GB per instance we are talking two arrays, or four disks even less worth the additional complexity.

    Then once you look at mature comercial inplementations and you start paying by the TB deduped it becomes utterly pointless. For sure an open source implentation can change that, but not one implemented in Java for crying out loud.

  3. Re:See also: LessFS on Open Source Deduplication For Linux With Opendedup · · Score: 1

    And given that it is not written in Java is likely to be much better performing.

  4. Re:How useful is this in realistic scenarios? on Open Source Deduplication For Linux With Opendedup · · Score: 1

    Yeah, but the problem there is the cost. We run on 17GB boot disks, so your 200VM's would require under 4TB of disk to store. I am sorry but 4TB of storage is peanuts and I can do that easily with a low end DS3400.

    Now the million dollar question to ask is how much does your dedupe solution cost? The reason being any dedupe that is supported against a virtualization solution we have looked at costs more than just buying the frigging disk. One then has to question the point of bothering with the extra layer of complexity.

    The level of dedupe in bulk storage is likely to be low as well, besides which the cost of dedupe on a couple hundred TB of disks is rediculas. Even for backup one has to wonder as well, tape is again really cheap, and dedupe for hundreds of TB is bloody expensive.

  5. Re:Stupid on Could Colorblindness Cure Be Morally Wrong? · · Score: 2, Insightful

    Indeed, that is fine. On the other side if a cure for deafness becomes available as a British tax payer who has to subsidize deaf children and adults through a range of help and other things given to them, I would consider it completely fair for these benefits to be removed should the cure be removed. So in this particular case should Molly's parents refuse the cure for themselves and Molly then they will have tens of thousands of pounds of annual assistance removed.

    Morally there is no reason whatsoever for me to pay to provide assistance for a condition that can be cured.

  6. Re:More like the Pot calling the kettle black on New Legislation Would Crack Down On Online Criminal Havens · · Score: 3, Insightful

    Screw that for years the USA harbored Irish terrorists. That is people convicted of blowing things up and murder. They did the same with north African terrorists that blew things up and murdered in France. Of course as soon as USA suffered a major foreign terrorist attack on it's own soil their tune changed.

    This double standard is why the USA has such a bad perception in most of the rest of the World.

  7. Re:Are you sure it's an IDE drive? More likely MFM on Need Help Salvaging Data From an Old Xenix System · · Score: 1

    You can still buy a brand new motherboard with an ISA slot today. Admittedly they are more expensive than a more mainstream motherboard, but anyone telling you that they don't exist is just plain ignorant. The ISA slot is not going to disappear anytime soon. If I have $100,000 upwards worth of equipment that is controlled by a computer using an ISA expansion card, I am not throwing it out because the computer is now kaput There are a whole range of USB to ISA and PCI to ISA converters that you can buy as well. Personally a new motherboard with ISA slot is the way I have gone in the past and would again.

  8. Re:Realistic Comparison - Tape vs Disk on Long-Term Storage of Moderately Large Datasets? · · Score: 1

    You didn't mention running costs. Tape cost almost nothing to run. I have big IBM TS3500 libraries at work with hundreds of tapes and lots of drives. The whole lot consumes less power than a couple of EXP810 shelves on the back of a DS4700/4800 and given the local climate could be stuck in a room with no air conditioning and still be fine. I could add in a new cabinet to the library with 1300 new tape slots and my power consumption would still be the same. With LTO4 at 1:1.5 compression that is another 2PB of storage.

    Compared to the equivalent amount of disk space I would be looking at thousands of pounds a year in electricity.

  9. Re:Big disks and ATAoE on Long-Term Storage of Moderately Large Datasets? · · Score: 1

    Clearly you have never used an enterprise grade tape solution have you. For large volumes of data tape is not only insainly cheap to purchase it is also insainly cheap to run as it consumes next to nothing in power and consequently needs very little if any cooling.

  10. Re:Another Option / Definition issues on Long-Term Storage of Moderately Large Datasets? · · Score: 1

    If you could not store 86TB of data without spending millions of dollars then you need to get out of the storage business as you have no clue. I look after over 300TB of spinning disk at work and we ain't spent millions on it as we don't have millions to begin with. For $60k I could put your 86TB on brand new un duplicated disk with brand new servers all on a five year maintenance contract.

  11. Re:Tape is your friend on Long-Term Storage of Moderately Large Datasets? · · Score: 1

    Yeah, but I would rather copy a LTO4 onto an LTO6 a couple of times than deal with hundreds of DVD's.

  12. Re:Tape is your friend on Long-Term Storage of Moderately Large Datasets? · · Score: 1

    With LTO3 and up you can go out and buy WORM tapes, that simply cannot be overwritten without hacking the firmware of a tape drive. With LTO4 and up you can have the tape drive encrypt it all as it is written to the tape.

  13. Re:Exactly. on Long-Term Storage of Moderately Large Datasets? · · Score: 1

    Hum, if you want support from a proper vendor, then for lots of disk we have found that the cheapest you can get is Dell's MD3000i which is rebadged LSI storage. We can get it full stacked out cheaper than a thumper from Sun. Sure it takes up more space than a thumper but it is cheap, takes a mix of SAS/SATA and does RAID6.

    Personally I prefer IBM's GPFS because it gives me ILM features. A whole bunch of big fat RAID6 arrays using 2TB SATA drives, some fast 600GB SAS 15k RPM drives. New data goes to the fast drives, and then gets migrated to the slower big fat SATA RAID6's. Throw in some 73GB 15k RPM SAS drives to hold metadata, and if you are using more than one storage array have your metadata replicated between them.

    Throw in some cheap servers as head nodes (the Dell R300 with redundant PSU/ boot disk and a DRAC card works well), with extra NIC's and an iSCSI card (take your pick between an expensive but easy to use Qlogic iSCSI HBA or a cheap and nasty Broadcom with iSCSI offload) and use clustered Samba to serve it all up.

    Need more space, add more arrays. Need more server bandwidth add more servers. One of your servers dies, not problem it is all clustered. Need to do an upgrade, it can all be done rolling.

    I only have 300TB of disk, but it is growing fast. Looking to put another 100TB in, in the next month or so.

  14. Re:Exactly what you're doing on Long-Term Storage of Moderately Large Datasets? · · Score: 1

    That depends on your definition of big.

    [root@nemo1 ~]# df -h
    Filesystem Size Used Avail Use% Mounted on /dev/mapper/VolGroup00-root
                                                7.8G 5.6G 1.9G 76% / /dev/sda1 251M 32M 206M 14% /boot
    tmpfs 7.9G 0 7.9G 0% /dev/shm /dev/gpfs 99T 81T 18T 82% /gpfs
    [root@nemo1 ~]# dsmdf /gpfs
    IBM Tivoli Storage Manager
    Command Line Space Management Client Interface
        Client Version 5, Release 5, Level 1.0
        Client date/time: 03/04/2010 08:39:45
    (c) Copyright by IBM Corporation and other(s) 1990, 2008. All Rights Reserved.

    HSM FS Mgrtd Pmgrtd Mgrtd Pmgrtd Unused Free
    Filesystem State Size Size Files Files Inodes Size /gpfs a 35.81T 457.44G 182560 843 6406389 17.85T

    The data is growing at an average of 1TB a week. Backing that up to more disk would really suck. Instead I have two tape libaries one onsite and one two miles down the road. It is backed up to both nightly.

  15. Re:Folks suggesting LTO-4 should also mention on Long-Term Storage of Moderately Large Datasets? · · Score: 0, Flamebait

    An folks that don;t know jack about LTO4 should shut the f$%& up. An LTO4 drive will adjust the tape speed so that it maintains streaming even if the data rate drops well below the max sustainable rate.

  16. Re: Will the Serial Console Ever Die? on Will the Serial Console Ever Die? · · Score: 1

    So what, what lights out management card does not have virtual CD/floppy/USB features?

    If you are serious about remote management you need proper lights out management, such as ILO, DRAC, RAS etc. depending on your server vendor.

    Put another way a serial console is not enough, you need to be able to give it a remote kick.

  17. Re:recent cellphone radiation reports on Killer Apartment Vs. Persistent Microwave Exposure? · · Score: 5, Insightful

    And then fail utterly to find a controlled study that shows repeatable results.

    Lets make this clear, in over fifty years of trying nobody and I repeat nobody has yet managed to do a REPEATABLE study that shows harmful effects of low level non-ionizing radiation.

    The key factor here is REPEATABLE. If it cannot be repeated it is just a meaningless statistical fluke.

  18. Re:It's this kind thing.. on Banks Accept Dubai Assassins' Stolen IDs · · Score: 1

    Still does not explain two of the alleged operatives fleeing to Iran. If that is accurate it would make Mossad involvement seem very unlikely, especially given that Hamas have plenty of other enemies.

  19. Re:Large sector size good? on Exploring Advanced Format Hard Drive Technology · · Score: 3, Informative

    IBM's GPFS is one, though it ain't free it does support Linux and Windows both mounting the same file system at the same time. They reckon the optimum block size for the file system is 1MB. I am not convinced of that myself, but always give my GPFS file systems 1MB block sizes.

    Then there is XFS that for small files will put the data in with the metadata to save space. However unless you have millions of files forget about it. With modern drive sizes the loss of space is not important. If you have millions of files stop using the file system as a database.

  20. Re:What the Judge Said... on Simon Singh To Appeal In UK Court Today · · Score: 3, Interesting

    If the British Chiropractic Association claim the use of chiropractic works for certain children’s ailments such as asthma, colic, and frequent ear infections, but refuse to provide any evidence that they do then one can only presume that they knew full well that that they don't work and are hence bogus.

    However this is immaterial to the case. In England and Wales (could be different in Scotland as we have a different legal system) the mere fact that Simon Singh's claim whether right or wrong was made and has damaged the claimant aka liabled him to the general public is all that is required in theory for the BCA to win. Truth is no defence against liable in England and Wales.

  21. Re:Missed out on Python on Learning Python, 4th Edition · · Score: 0

    Yes because no matter what language I use I write readable code - PERIOD. Readable code has nothing whatsoever to do with the language. I can write unreadable Perl and write readable Perl. I can write readable Python and unreadable Python. Well I could if I new Python :-)

  22. Re:Too wordy on Learning Python, 4th Edition · · Score: 1

    The copy on my shelf says 272

  23. Re:Move to Canada on Health Insurance When Leaving the Corporate World? · · Score: 1

    That is around 9700GBP, or close to my combined income tax and national insurance bill and I earn getting on for double medium income in the UK. My quality of life would truly suck if I had to pay that much for health care.

  24. Re:There's more to this story on Our Low-Tech Tax Code · · Score: 1

    Where did you get that shit? You will be telling me next that the NHS would kill of Stephen Hawking.

  25. Re:There's more to this story on Our Low-Tech Tax Code · · Score: 1

    Except health care outcomes in the USA for *ANY* segment of the population, and it does not matter how you slice it are *WORSE* than the corresponding segment in the UK. Now I am not saying that the NHS is perfect because it is not, but that is a horrific indictment of the US system, given the amount that is spent in the US on health care and the fact that millions of people have no cover.

    The curious thing is that by any objective measure of health outcomes the USA health care system is very poor, in comparison to any other western industrialized nation. Yet despite this many if not most Americans live in the deluded belief that theirs is the best health care system in the world.

    I also remember when Clinton was trying to reform health care reading that the US spent more pushing paper about working out bills than the UK spent in total!! The USA system is just broken beyond belief.