3/29    A friend wants a PC (no mac). She doesn't want Dell. Is there a
        good place that can custom build for you (SSD, large RAM, cheap video
        card--no game)?
        \_ As a side note: back in my Cal days more than two decades ago when
           having a 387SX made me the only person with floating-point hardware,
           most machines were custom built.
           most machines were custom built.  -- yuen
           \_ did that make you a better programmer/gamer/etc?
              \_ No.  I was writing a pre-emptive multi-threading library as
                 part of an effort to port some floating-point-heavy Unix code
                 to DOS for CS199.  I couldn't figure out how to save the
                 "context" of Borland C++'s floating-point emulation library.
                 The options were to either mask context switching around all
                 floating point operation or get a co-processor.  I ended up
                 getting the co-procoessor.  -- yuen
                 to DOS for CS199 (http://ssli.ee.washington.edu/~bilmes
                 mypubs/bilmes1992-icmcmood.pdf).  I couldn't figure out how
                 to save the "context" of Borland C++'s floating-point
                 emulation library.  The options were to either mask context
                 switching around all floating point operations or get a
                 co-processor.  I ended up getting the co-procoessor.  -- yuen
        \_ Why not Dell? What is it about Dell she doesn't like?
        \_ Fry's? HP? Lenovo? The only place I have done this is Lenovo.

1/4     I want to test how my servers behave during a disk failure and
        a RAID reconstruction so I want to simulate a hardware failure.
        How can I do this in Linux without having to physically pull
        a drive? These disks are behind a RAID card and run Linux. -ausman
        \_ According to the Linux RAID wiki, you might be able to use mdadm
           to do this with something like the following:
              mdadm --manage --set-faulty /dev/md1 /dev/[disk]
           link:preview.tinyurl.com/88vwcvu [raid.wiki.kernel.org]
           [ Maybe this only works for software raid - I don't have much
             experience with hw raid on linux ]

11/27   CalMail has been down for a few days (hardware failure and database
        corruption -- sounds like fun!) and is starting to come back online.
        Looks like they're planning to outsource all campus mail to either
        Google Apps or Microsoft 365 as part of Operational Excellence.
        <DEAD>kb.berkeley.edu/jivekb/entry!default.jspa?externalID=2915<DEAD>
        \_ http://ist.berkeley.edu/ciocalmailupdates/november-30-2011

9/13    Thanks to Jordan, our disk server is no longer virtualized. Our long
        nightmare of poor IO performance should hopefully be over. Prepare for
        another long nightmare of poor hardware reliability!
        ...
        Just kidding! (I hope)
        In any case, this means that cooler was taken out back and shot, and
        replaced with Keg, a real machine with real disks. Right now it's not
        running at 100%, but already you should notice that soda's not only
        fast, it's a fucking miracle compared to the past few years. I
        personally blame unforeseen edge cases in a poisonous combination of
        ZFS+NFS+OpenSolaris+1000s of users with a system too big to fail.
        Indeed, syncing the data away from cooler took two continuous weeks.
        It's no wonder it's taken until now for a very capable VP to be up to
        the task of partially unbreaking the setup. Note - we no longer have
        any VMs running off of virtualized disks stored on a NFS mounted disk
        which, itself, was virtualized. Hmmmmmmmm. Though those were mostly
        useless VMs you never saw. :P

        So anyways, as mentioned earlier, Keg isn't at 100%, but it's up. It
        looks good enough to keep for a bit, but it originally had a bunch of
        Raptors or some such. The disks are still there, but the RAID cards are
        most likely broken. We'll leave it to jordan to evaluate the server
        needs and fix accordingly. As it is, RAIDing fifteenish 10000RPM disks
        so you can edit motd SUPER-EXTRA FAST!!! is probably not a great use of
        time. We'll see where our less-shaky infrastructure takes us in the
        future. --toulouse
        \_ cooler is dead. Long live KEG!
        \_ Good work guys, thanks! #1 lesson here: don't virtualize i/o
        \_ Good work guys, thanks! #1 lesson here: don't virtualize disk i/o
           intensive applications. -ausman
           \_ That is a good lesson but definitely not the #1 lesson.
              * Exporting thousands of filesystems: bad idea, no matter how
                easy it makes backups and ZFS snapshotting.
              * Using an OS with superior filesystem support is a bad long-term
                solution if nobody but the original installer knows almost
                anything about it
              * Choosing ZFS...the jury's still out.
              * Maintaining FreeBSD 7 and 8 and OpenSolaris and Debian...kinda
                hard.
              * All of this, on top of virtualized disk i/o - bad news.
              \_ Even after I collapsed NFS down to one filesystem, when our
                 FreeBSD boxes came back online and started automounting
                 thousands of filesystems apiece, the NFS server again ground
                 to a screeching halt (taking soda and friends with it).
                 Switching to one /home mount per server restored NFS's
                 snappiness; I suspect that even a virtualized NFS server could
                 perform well without the filesystem woes.  --jordan

7/1     Does anyone have good info, either published or anecdotal, about
        SSD failure rates. I am seeing an alaarmingly high failure rate
        in Enterprise applications at scale. -anonymous coward
        \_ http://www.codinghorror.com/blog/2011/05/the-hot-crazy-solid-state-drive-scale.html
           http://www.pcworld.com/businesscenter/article/213442/solid_state_drives_no_better_than_others_survey_says.html
           http://www.hardware.fr/articles/810-6/taux-pannes-composants.html
           Annecdotal: yes, HD-to-HD it's comparable, if it is of high end
           brand. HD-to-HD, SSDs suck if it is OCZ and cheap brands. Byte-for
           -byte, it is horrible. My Samsung failed catastrophically in 11
           months but was perfect before that.

8/19    Anyone have a good MLC vs. SLC writeup that is less than a year
        old? I am investigating making a big SSD order. -ausman
        \_ how much performance gain are you expecting from using SSD?
           Let's say your 7200RPM yields 70 MB/sec (sustained) read and
           your new SSD yields 160 MB/sec read, and that your application
           is mostly streaming media (so the platter is continuous read,
           mostly). Do you expect a 2X performance? Now, let's say your
           application is mostly short random meta-data reads and your
           7200RPM now only yields 32MB/sec read, while your SSD still
           yields 160MB/sec read. Do you expect a 5X performance in your
           response time or throughput, or both? I'm quizzing your ability
           to perform calculations as an architect.

7/22    Playing with dd if=/dev/random of=/dev/<disk> on linux and bsd:
        2 questions, on linux when <disk>==hda it always gives me this off
        by one report i.e. Records out == records in-1 and says there is an
        error. Has anyone else seen this?  Second, when trying to repeat this
        on bsd, <disk>==rwd0 now, to my surprise, using the install disk and
        selecting (S)hell, when I try to dd a 40 gig disk it says "409 records
        in, 409 records out, / full".  Which is far well below what I'd expect
        on a 40 gig.  If i actually fdisk the drive, and put a partition  I can
        dd if=/dev/random of=/dev/[rwd0a],[wd0a] but still get the off by one.
        Has anyone else seen this weird behaviour?  I'm pretty sure that I
        should be able to dd to rwd0 in bsd just as I can dd to hda in linux.
        This is an IDE drive btw.   Also I set bs=512 and bs=8b to try it out
        multiple ways.

5/22    Duke turns of its Usenet server:
        http://news.duke.edu/2010/05/usenet.html
        \_ Imminent death of the net predicted.
        \_ :( Sucks. I wish Berkeley hadn't turned off Usenet, either.

2/4     My external hard drive (not used) is starting to emit
        a ticking sound every 5-10 seconds even though I'm not
        reading or writing anything. It used to be quiet. Is
        this a sign that it's about to fail?
        \_ would have to hear it to be sure, but it certainly
           is likely.
        \_ nothing in the bg doing anything? is this while plugged into the
           computer or not?
           \_ it's doing this even though I'm not mounted, so
              I'm getting a bit worried. It never did this before.
              BTW it's a WD-1TB green drive.
              \_ check SMART status?

10/27   I just read an article that Facebook had moved their database
        to all SSD to speed throughput, but now I can't find it. Has
        anyone else seen this? Any experience with doing this? -ausman
        \_ I hope you're not running mission critical data:
           http://ask.slashdot.org/story/09/10/27/1559248/Reliability-of-PC-Flash-SSDs?from=rss
        \_ Do you have any idea how much storage space is used by Facebook,
           and what the cost implication would be to move *ALL* the data
           to SSD? I believe they may have experimented with using SSD
           as a 3rd tier cache layer between RAM and disk but the cost
           of *ALL* data is simply prohibitive.                 -kchang
           \_ SSD is $3k/TB now, I doubt that Facebook has more than 1PB of
              total data, so that would only be $3M. They probably spend more
              than that on electricity every year.
        \_ Are you thinking of MySpace, perhaps?
           http://www.theregister.co.uk/2009/10/14/myspace_fusionio
           It's not clear from that article whether they're using flash for
           all their storage, or just for caches.
           \_ Yes, that is it, thanks. -ausman
           \_ SSD doesn't make any sense in terms of $ for stuff like video
              and music and pictures. It does make sense for frequently used
              metadata. The question one should ask is, what is the infrastructure
              of MySpace like, is it using sharded MySQL or something else?
           \- our 12x or 16x RAID is faster than our SSD [high $$$ SSD, medium
              quality RAID. These are generally large seq reads, not small
              random ones.].
              \_ How fast is each?
        \_ Just because SSD r/w is 4X faster, doesn't mean your system will
           run in 1/4 the time. You gotta take in account of request overhead,
           and processing time (complex MySQL join is particularly expensive).
           My friends in the SSD industry said basically that speed-up wasn't as
           mind-blowing as they originally anticipated, and that if the
           application isn't SSD tuned, you may not get the amazing speed-up
           that you thought you would get.
           \_ I know that some FS are not well designed to take advantage of
              SSD, what other tuning to I need to be aware of?
              \- our biggest speedup was buying a lot of memory [+100gb]
                 and judicious use of compression ... with lots of cores
                 but limited bus IO, this is generally a big win even using
                 gzip style compression, although there are some faster
                 compression systems which dont compress as much in space
                 but are like 4x faster than gzip. BTW, does anybody other
                 than GOOG use "zippy"? Are there any tools/filesystems which
                 use zippy, or is that GOOG intenal. (we didnt spend much time
                 researching this ... we want to throw money and a little time
                 and the problems to mitigate it ... at some point we'll start
                 indexing, which is what will get us to the orders-of-mag
                 improvements).
                 \_ I think Zippy is fine tuned for Bigtable data, which is
                    basically value-key-time. Since they know their input
                    well, you don't need an all purpose gzip, hence the
                    10X encode rate and 3X decode rate over gzip. By the way
                    if you find Zippy implementation, let us know!
                    http://feedblog.org/2008/10/12/google-bigtable-compression-zippy-and-bmdiff

7/28    Does it make sense to defragment disks on VMWare? My 80GB disk
        on VMWare isn't really using 80GB, it just uses what it needs.
        Will defragment do anything to it?
        \_ If you want to speed up disk operation in your VM, it's best to
           defragment the disks in your VM, then defragment the disk on your
           host machine where the VM files are.
           If you want to shrink the VM files, you need to defragment the disks
           in your VM, then do the shrink operation on your host machine.
           \_ I'm running WinXP under Mac's VMWare Fusion. I can't defrag
              Mac's file system. So is defragging useless under Fusion?
              \_ The mac automagically defragments for you.
                 \_ Don't question the Steve. Everything from Apple is good.

5/10    Is it possible to install MacOS on a built yourself PC?
        \_ There are some hacks which should make it possible.
           http://osx86.thefreesuite.com.  It's a license violation,
           of course.  -tom
        \_ One Step Install:
           http://dailyapps.net/2007/12/hack-attack-install-leopard-on-your-pc-in-just-one-step

5/10  I have large spare disks on several machines, all on same net.
      How can I combine them into one big visible network file system?
      \_ you want either a clustered file system (eww expensive) or look into
         smoething like a distributed file system (afs, global fs).
      \_ The startup I work for makes a distributed NAS product that's free
         for the first four terabytes if you'd like to try it.
         http://www.parascale.com                      --alawrenc

5/2     I have a 1Gbps switch. It is connected to a SANS RAID-1 with
        two WD Green 1T that takes 1Gbps connection, as well as a laptop
        that is also 1Gbps. Online benchmarks show *sustained* transfer
        rate of about 1/2 of 1Gbps with the WD Green... cool!!! But
        in practice, I'm only getting 68Mbps (read SANS->PC file transfer),
        far below what I expected. The switch says both devices are using
        1Gbps. BTW, setting the SANS on Jumble Frame/no jumble doesn't
        change the transfer rate at all. What's going on?
        \_ Are you limited by your PC's drive speed? 68 Mbps is 8.5 MB/s
           which may be all your drive can do. What do you get copying
           data on your drive locally?
           \_ When I copy a 796MB file from the same drive, it took about
              a minute. So I figured... 796MB read and 796MB write (gross
              56 sec. So I figured... 796MB read and 796MB write (gross
              assumption that they're about the same, but false).
              That's 796*2 / 60 = 26MB/sec. So I hardly think I'm limited
              by my drive.
              That's 796*2 / 56 = 28MB/sec. So I hardly think I'm limited
              by my drive. I'm almost certain the bottleneck is due to
              something stupid I did, but seriously, I checked that NAS
              is in fact set to use 1000Mbps (or no connection), and I know
              my laptop transfer speed is super fast at work. Reallly baffled...
              my laptop transfer speed is super fast at work. Reallly
              baffled...
              \_ Your drive can't do any better than 224 Mbps. Less over
                 the network given the overhead. Figure maybe 0.66 of 224
                 for 135 Mbps. That's only a factor of two different from
                 what you are seeing. Maybe there is some caching going on
                 to give you that 2x. If the NAS is *really* capable of
                 500 Mbps (which is what my $100K Netapp striped across
                 many disks can do so it seems unlikely) then I'd look to
                 your drive or your network card. Could also be that your
                 switch sucks. Try to connect with a crossover cable to
                 see if that helps.
        \_ What is the Sustained Transfer Rate for the disks in your NAS?
           Odd are it is about 25MBps/per disk. How wide is your stripe?
           http://www.storagereview.com/map/lm.cgi/str

4/28    Anyone here have experience with solid state drives in the server
        environment? I am mostly worried about reliability, longevity, and
        performance. I've read a couple of articles saying that OSes
        really need to be SSD optimized before this can be considered
        truly viable and yet every vendor from Dell to Sun to Apple is
        shipping these. Are they ready for prime-time on an important
        production server?
        \_ Theres been a lot fo advances in SSD recently toward load levelling
           (to spread the erase cycle wear) an overall endurance.  They're
           actually a viable (if expensive) option now.   Theres been some good
           work looking at using them for ZFS logs for zvols involving lots
           of slower storage.
                If you're looking at using htem now they'll probably be
           replaced due to obsolescence before they 'go bad'.   -ERic
           \_ I'd still be a bit wary of using them for something like a
              high write db server, but yeah for user stuff go for it.

4/8     I'm fed up with having up download iTunes every month or so.
        Frigging annoying. WHY do I have to upgrade?
        \_ Apple hasn't figured out how to fully take over all of your
           spare cycles with their autoappleupdatesoftwaredaemons.
        \_ Well, you don't *HAVE* to update do you?
           \_ That's right. The latest update is to provide support for
              HD Movie Rentals. I didn't upgrade. --iTunes Employee
              HD Movie Rentals. I didn't upgrade.
        \_ They have to make sure my java version doesn't work.

4/5     Tuesday at 10:00PM I will be taking down Soda temporarily to migrate
        disk arrays.  This is an expected part of our move to the new server.
        The downtime should be minimal - I'll be remounting /home read-only
        for an hour or so while I run a final mirror over to the new array
        and then will reboot Soda onto the new array. The final migration
        will come at a slightly later time - this is in preparation for
        moving mail handling.

        We apologize for the inconvenience.

3/30    CF/SD Disk database:
        http://robgalbraith.com/bins/multi_page.asp?cid=6007

3/19    Pres. Obama's gift of DVDs doesn't even play in England
        http://www.telegraph.co.uk/news/newstopics/mandrake/5011941/Gordon-Brown-is-frustrated-by-Psycho-in-No-10.html
        \_ maybe it's a crypto-protest against region'd DVDs

2/16    What's the best way to turn your iPhone into a hard-drive/backup
        device (while not jail-breaking it)?
        \_ I use a program called Discover which turns your iPhone into a FTP
           site using wifi.  -scottyg
           \_ Looks like it's free, great! Any problems you've seen on it?
              Some people online says there are problems with big files.

2/15    How many watts does a typical desktop (not counting monitor)
        use when it is completely idling (top shows at most 1-2% CPU),
        assuming normal SATA HD, video card, 2-4G RAM?
        \_ Google did a published study where they determined that the
           largest consumer of energy in a computer is the power supply
           itself. Figure that if you have a 400W power supply you are
           using at least half that if not more even when idle.
           \_ that totally blows. Laptop #1!!!
              \_ Power supply manufacturers need an incentive to make more
                 efficient power supplies. It's probably not hard to do.
                 Google was talking about pushing manufacturers on that
                 issue, but one company is probably not enough.

1/26    Has anyone tried using solid state hard drive as the OS+swap to
        improve performance?
        \_ you know I haven't seen my machine touch swap in years.  How
           much RAM do you have?
        \_ Set it as swap and it might last a few hours.  You know SSDs have a
           write limit, right?
           \_ "I did not know that." How many times can I write to my
              flash drive before kaput?
           \_ All drives have a write limit.  10-15 years ago flash had a
              write limit in the 1000's of writes per sector.  Those days
              are long gone and SSDs a simaler level of mean time before
              failure as hard drives.
           \_ "Industrial Strength" CF cards (used in Industrial computers
               / automation controllers) has a guarentee life of about
               100k write.   Most of the SSD in you use are *NOT* industrial
               grade by a long shot, as industrial grade is *A LOT MORE*
               (e.g. 4x or more) expensive.

1/14    http://www.amazon.com/Appetite-Self-Destruction-Spectacular-Industry-Digital/dp/1416552154
        \_ where's that gonna go after
           http://news.cnet.com/8301-13578_3-10133425-38.html

1/11    "quota -v" doesn't work anymore?
        \_ I was told on motd a few years ago that soda no longer had disk
           quota.
           \_ We don't, but reserve the right to slap hosers, since disk space
              isn't unlimited. --t

12/31   SOMEONE PLEASE FIX KEG!  Why does it keep crashing?
        \_ It's hard to tell. The risk of crashing seems to increase with
           disk activity, or so it seems, but debugging the kernel doesn't
           seem to yield much info. Email root for detailed info; I don't
           have a deep understanding of the problem --t
           \_ How about a backup LDAP server? I don't need files as much
              as access.
              \_ We do (did?) have a backup LDAP server. I think that it's also
                 suffering from old failing hardware, but again, root
                 (particularly Steven) hould have accurate-up-to-date info. The
                 new server won't arrive for a little while yet, and it's only
                 one server to replace several servers that have been failing in
                 recent times.
                 one server to replace several servers that have been failing
                 in recent times.
                 $PLEA_FOR_DONATIONS
                 --t
                 \_ Where do we send money to again? -ausman
                    \_ A check made out to the CSUA and mailed to 337 Soda
                       Hall, UC Berkeley, Berkeley CA 94720 or PayPal to
                       politburo@csua.berkeley.edu are the preferred methods.
                       --t
              \_ Okay, I got some details on server roles, for reference.
                 soda: login, www, mysql
                 scotch: DNS
                 vermouth: mail, primary LDAP
                 keg: NFS, secondary LDAP, backups
                 screwdriver: linux mirror, mysql
                 lifesaver (now dead): LDAP, backups
                 ---
                 Screwdriver and scotch don't mount NFS, so generally are the
                 only computers to survive a keg failure. --t
              \_ New hardware is on the way!  ETA 1 wk or so
                 Then give us a week to set it up and whatnot

12/15   I have a few 100G and 200G drives that I want to dump. Do people
        actually buy used legacy HD on eBay?
        \_ Yup they do -- researchers doing articles on the lack of safety
           of data on computers often go to eBay, buy used hard drives,
           and mine all sorts of goodies from the not, or poorly, wiped
           drives.

12/9    Another idea for the CSUA that lets you spend money and maybe get some
    cool toys. Instead of buying a beefy server (like say, a massive server
    with 20 386DX processors), buy a few cheap machines (like the ones
    mentioned below) that have good disks and work on failover / load
    balancing. A netscaler or other piece of hardware is complete overkill,
    but maybe hacking an OpenBSD box could do the trick. The idea is that
    this project:
    - Lets you learn a lot about failover/redundancy/scalability, which
      are real-world, relevant skills and also interesting. The fake
      CSUA head-hunter would be all over these real skills.
    - Increases the reliability/performance of soda / csua machines.
    - Gives a justification for buying new hardware.
    - Give a cause for the CSUA.
    Good idea: ..
    Bad idea:
        \_ 386DX is cool: http://csua.com/?entry=18447
        \_ Too late, the server has already been ordered. We already have cheap
           machines, and we're note exactly getting pounded on the web, nor do
           we have sites to scale up. We already have stuff on FreeBSD as well
           as backup LDAP. The point of failure is keg and the fact that we have
           16 small disks. We've also ordered new disks to rectify that. I gat
           16 small disks. We've also ordered new disks to rectify that. I got
           the idea that you'd rather we work around a lack of resources than
           exploit more than we think we need...or at least, that's what I keep
           hearing. I don't think that's really going to help. --t
           as backup LDAP. The point of failure is keg and the fact that we
           have 16 small disks. We've also ordered new disks to rectify that.
           I got the idea that you'd rather we work around a lack of resources
           than exploit more than we think we need...or at least, that's what
           I keep hearing. I don't think that's really going to help. --t
           \_ Good decision. Especially good to just make a decision and
              start fixing things, instead of debating endlessly.
           \_ I like what you did more than my proposal! Stability, not speed,
              should be the first factor. Good disks and stable builds are
              two of the biggest contributors to stability. Do you still hit
              up alums for donations? -op
              \_ We were going to hit up alumni for donations for the server
                 before we got it, but the stability was (literally) getting us
                 down and so we decided to get the new server. Steven, last I
                 heard, plans on installing ESXi (the free one) on it setting
                 it up. It should be in place by mid-to-late-January, possibly
                 heard, plans on installing ESXi (the free one) on it setting it
                 up. It should be in place by mid-to-late-January, possibly
                 earlier. --t

12/5    What would you guys think of this?
        TYAN Tank barebone
        http://tyan.com/product_barebones_detail.aspx?pid=353
        2x Intel Quad Xeon E5420
        http://www.newegg.com/Product/Product.aspx?Item=N82E16819117147
        16GB ram
        http://www.newegg.com/Product/Product.aspx?Item=N82E16820134633
        SAS hard drives
        http://www.newegg.com/Product/Product.aspx?Item=N82E16822116058

        Total cost <$3000, and it seems like this would be an ESX supported
        configuration if we ended up going that way (and there's no reason
        that it would preclude any other virtualization option either)

        The goal here would be to eliminate the other servers in various states
        of 'broken' (scotch, screwdriver, lifesaver) and replace it with this
        one.  Once it was up and tested, we'd migrate soda there and then reuse
        the hardware from Soda for something else.  Thoughts?
        \_ Why do you believe soda would not run on a $1-1.5k machine:
           3gz proc + 1tb disk + ~4gb memory. I suppose something on the
           higher end of this range with two slower processors might be
           better than a 1proc. I'm guessing you'ld be better off with
           one two lower end machine. Especially if you want one to play
           around on.
           \_ He's talking about replacing 4 servers with 1 server, not
              just replacing soda.  Soda is already way overpowered for
              what it's doing right now.  It doesn't need to be a dedicated
              machine anymore.
              After that I really can't tell what you're trying to say.

12/5    In RAID 1, when I write, is it a synchronous write to both drives?
        In another word, is my write complete upon verifying BOTH drives
        have written (indicating the slowest drive = bottleneck)? How about
        read? Does RAID 1 require reading both drives, comparing results,
        then returning them? Or can it do striping read?
        \_ Write completes when it has written to both so RAID1 writes are
           usually a little slower than writing to a single disk. Reads are
           from one only, which is why RAID1 has very good read performance.
           You can read different data off of both drives at the same time.
           Remember, RAID1 doesn't stripe at all. Your drives should be
           identical to avoid one being much slower than the other.
           \_ What if the first read gives me bogus data (or data that
              is different than the second data because a neutrino particle
              changed a few bits)?
              \_ You don't use RAID1 to protect against that. If the two
                 bits were different how would the system know which was
                 correct? There's no parity information with RAID1.
              \- hello, YMWTGF(zfs, merkle tree). btw, if you see this happen
                 i'd recommend focusing on the neutrino changing some bits
                 and forget about the integrity of the data. --psb
        \_ We should start up that Silicon Valley CSUA blacklist again, since
           CSUA has a hammerlock on the economy.

12/4    A question to you old crufy alumni: So lately we've suggested
        VMs, and been asked why it's necessary. We've suggested top-of-the-line
        hardware and been told we don't need that much power. So I'd like to
        ask -- what exactly do you think the CSUA is supposed to _be_?
        \_ Noone said VMs weren't needed.  They suggested you use the
        \_ No one said VMs weren't needed.  They suggested you use the
           there is quality, free VM software out there that you could
           use.  Asking for the for pay stuff is fine, but you aren't
           likely to actually get it, so take what's free and works.
           \_ I guess...would it hurt that much to ask for the expensive stuff,
              though?
              \_ When the company supplies a free version that works quite
                 well you will get a lot more traction towards getting the
                 expensive version if you have used the free version, shown
                 that it's actually getting some use beyond "we installed
                 it cause it was kewl" and that there would actually be
                 some advantage to having the pay version.  VM stuff IS
                 really cool and worth playing with.  The free version
                 exists because really, until you have a large install
                 there's really no need for expensive management tools.
        \_ Offering a rant, then an answer for my own question:
           Sure, VMs aren't strictly necessary, or people can go to
           OCF if they want access to computing resources or practice
           with sysadminning. But they'd be supported by OCF, which
           has basically nothing to do with the CSUA (anymore), not
           to mention they're bound by whatever the OCF is willing to
           allow/install/support on their servers. VMs let us tinker w/o
           worrying about giving root, or a bad program taking down the
           server or monopolizing server resources. As a consequence
           it is also useful for stability's sake, independently of
           offering students VMs, as we can separate services out and
           move them between servers as load necessitates. Do we need
           the speed? We'd certainly see some benefit. Mail processing
           is pretty laggy, and even with my donation of vermouth (with an
           Athlon 64 2800+; it replaced a 500Mhz computer) it struggles
           to keep up. Upgrading away from our failing 16-disk ~5TB array
           to something newer and less failure-prone may not be manly -
           but it certainly seems sane to me. If we're going to spend
           $2k to replace our servers, why not spend $3k to replace our
           servers and move the CSUA forward?

           I doubt anyone objects to getting equipment so the servers
           suck less, but I think that in order to keep the CSUA from
           stagnating, we need to seriously move forward in what we offer
           students. Attendance to helpsessions isn't what it once was,
           and the computing services the CSUA currently offers simply
           aren't as attractive to students anymore. Anyone can find a
           UNIX login server offered by the campus with webspace (w/
           CGI) and shell access. Most people now join the CSUA as a
           social group or to use the office machines and hang out. At
           this rate we're going to lose our technical relevance,
           and what then? Regardless of overlap with the OCF, we
           need to move our technical resources forward and attract
           technically-minded students again, because at the moment,
           we're not. The CSUA is wandering at the moment, without a
           clearly-defined direction. If you guys are suggesting that we
           don't need to do more for students in this area, then I'd like
           to hear your thoughts on what we do (other than the obvious
           'build better student-alumni relations' line).
           \_ I dunno, the CSUA hasn't had a clear direction since I joined
              in '98.  (Web and email hosting wasn't terribly relevent then
              either.) We started mentoring in 2002, but I it always seemed
              either.) We started mentoring in 2002, but it always seemed
              like a bit of a flop to me.  I've noticed video games have been
              more important since I left.  Anyway, the CSUA is always trying
              new things.  I don't think that's a bad thing.  I just don't
              know if we need an i7 fileserver.
              know if we need an i7 fileserver.  I guess I don't really see
              how improving the HW -> more members.  I don't mind, really,
              I'd like soda to work better, but I don't know who's going to
              join because we have a nice fileserver or VMs for sys admining.
              It's not a bad thing, but it's all stuff people can do at home
              now.  It's good for current members, but I don't really see it
              as attractive to new members.
           \_ I already talked about this down below.  The primary function
           of the CSUA is to be a social, community-building organization.
           The machines exist to support that function.  It used to be that
           the services the CSUA provided on its hardware were directly
           contributing to the community--there was a time when having an
           email server and web hosting was a really cool thing, and people
           would join the CSUA just for that.  Then they'd get sucked into
           using wall and MOTD and become part of the community.  But now
           most of the services the CSUA historically provided are no longer
           interesting; everyone's got more email addresses than they want,
           web hosting is free and better elsewhere, and wall and MOTD are
           dead.  Providing virtual machines isn't an inherently bad idea,
           but what's the upshot?  How will you use it to help build
           community?  What services will you provide that contribute to
           community?  -tom
              I guess I don't really see how improving the HW -> more members.
              I don't mind, really, I'd like soda to work better, but I don't
              know who's going to join because we have a nice fileserver or VMs
              for sys admining. It's not a bad thing, but it's all stuff people
              can do at home now.  It may be good for current members, which is
              fine, but I don't really see it as attractive to new members.
           \_ I already talked about this down below.  The primary
              function of the CSUA is to be a social,
              community-building organization.  The machines exist to
              support that function.  It used to be that the services
              the CSUA provided on its hardware were directly
              contributing to the community--there was a time when
              having an email server and web hosting was a really cool
              thing, and people would join the CSUA just for that.
              Then they'd get sucked into using wall and MOTD and
              become part of the community.  But now most of the
              services the CSUA historically provided are no longer
              interesting; everyone's got more email addresses than
              they want, web hosting is free and better elsewhere, and
              wall and MOTD are dead.  Providing virtual machines
              isn't an inherently bad idea, but what's the upshot?
              How will you use it to help build community?  What
              services will you provide that contribute to community?
              -tom
              \_ perhaps having VMs available would make more interesting
                 helpsessions feasible, as we could go in-depth on some topic
                 and provide an avenue for students to go further in-depth
                 after the fact (i.e. helpsession setting up X piece of cool
                 software and giving them a sandbox to play in). --toulouse
        \_ As a crufty alum, I think the CSUA should encourage undergrads
           to learn about software, cs, computing, &c. by providing ugs
           with opportunities to try out and develop new things in a friendly
           setting.
           When I was an ug there were several CSUA projects, such as the
           lottery scheduler, that offered such opportunities.  I think
           that the VM proposal would offer a similar opportunity and should
           be pursued.
           Re VM software - I agree with a pp that setting up a free sol'n,
           such as Sun's virtual box, is probably a better idea than asking
           for a VMWare donation. IMO, the process of learning how to make
           free software work for one's needs teaches one about much more
           about software and computing than installing and configuring a
           pre-packaged commercial sol'n does.
           Re HW - While I understand the attraction of getting the latest,
           1337est hardware, my experience suggests that going with a sol'n
           of the previous gen hardware is always preferable in terms of
           maintainability and reliability.
           maintainability and reliability. -crufty-alum
           \_ :| Well, the Core i7 Xeons as I understand it will be coming out
              some time after the Core i7 has been out, so perhaps that will
              positively affect their quality? I see what you're saying, but
              I've never really been burned by new hardware, other than lack
              of drivers and the like. Care to share some stories? --toulouse
              \_ Well, mostly the problems I've had are with the drivers for
                 the motherboards that are required to run the most recent
                 cpus.  We had no end of problems with an Intel MB for the
                 first round of Xeon chips (there were no reliable linux
                 drivers for the onboard gig-e ethernet, the sata controller,
                 the scsi controller, or the lights out management module).
                 I had similar problems with nforce boards for AMD chips and
                 recall similar problems when the PPros and P2s first came
                 out as well.
                 For the record, I'm not opposed to the CSUA buying the latest
                 hardware.  I'm just not convinced that the projects you want
                 to do (e.g. VM) require it, i.e., what is the difference btwn
                 getting a 4 core C2D system (or 2 C2D systems) and getting one
                 Core i7 system?  [If there is a big difference, I apologize
                 for my ignorance - I am not an engineer anymore and have lost
                 touch with the details of recent cpu performance] -crufty-alum
                 \_ It seems practicality and wisdom are prevailing over our
                    nerd-rections, and we may go with a Dual C2Q Xeon server. We
                    await final word from the pupp--er, President. While we're
                    at it, we'll probably get new, better hard drives.
                    nerd-rections, and we may go with a Dual C2Q Xeon server.
                    We await final word from the pupp--er, President. While
                    we're at it, we'll probably get new, better hard drives.
                    --toulouse

11/13   lolz, do the cops delete their copy of your data when they return
        your computer?
        http://berkeleydailyplanet.com/issue/2008-11-13/article/31568?headline=Long-Haul-Gets-Computers-Back-Wants-UC-to-Delete-Seized-Info
        http://tinyurl.com/6hfntp
        \_ Did your 22 yo g/f dump you or something? If you keep deleting
           my comments, I am going to start deleting yours, btw.
           \_ lulz

11/11   I want a home nas so my laptops aren't always running out of space,
        what's a good buy?
        \_ I did a bunch of research and ended up with Dlink DNS-321. It
           is newer than DNS-323 with less problems. It doesn't have as
           much buffer as DNS-323, but if you're transfer files bigger
           than 32 megs, the sustained rate for both are equivalent. I
           bought a Western Digital green (5W 5400RPM) and mixed it up
           with Seagate Barracrasha (9W 7200RPM) for Raid-0. I know it's
           not the best of mix (5400 and 7200) but I did it for redundancy
           reasons. I don't trust either brands as I've had failures in the
           past from these two companies. Hopefully, the mix will allow
           one crash a lot sooner than the other, instead of both at the
           same time. If performance is a big thing for you, get two
           7200RPMs. By the way, WD green is really quiet!!! I love it,
           I can barely hear it at night. Seagate is slightly noiser, but
           not by much.
           \_ Can it hotswap disks?
              \_ Not this one, but DNS-34X something can do it (the one
                 with 4 drives). You'll pay a premium for it though. I got
                 DNS-321 because it's quite good for the price.
        \_ You can get a used NetApp on EBay for really cheap.
        \_ I'm told that Drobo + DroboShare is very nice.  I also know
           a several people who are very happy with Apple's Time Machine.
           \_ drobo sucks. it's loud and doesn't save your data. droboshare
              doesn't like non-ascii characters, even if it is HFS+J formatted
              \_ Any opinions on the netgear nases?
        \_ Qnap.  Anyone heard of it?

10/27   What's a good RAID-1 SATA card to get for the PC?

8/27    always have offsite data storage
        http://www.berkeleydailyplanet.com/issue/2008-08-21/article/30943?headline=Campus-Police-Raid-Long-Haul-Seize-Computers-Disks-Drives
        \_ This article also has a lot of funny lines in it.  Ah Berkeley.
        \_ This is on concern I have about co-los and cloud computing. What
        \_ This is one concern I have about co-los and cloud computing. What
           happens when the FBI raids Amazon S3 for having child porn on the
           servers? I think the people who say that the future of computing is
           a cloud somewhere run by an outsourced IT department are on crack.
           I'd never run any business I cared about in that manner.
           \_ read up about safe harbor
              \_ Why is safe harbor relevant?
        \_ When I was a POW, we didn't HAVE offsite storage!
           \_ I dunn, being a POW is kinda like being in offsite storage...
           \_ The FBI is not stupid. They are not going to shut down a bunch
              of legitimate companies just to catch a few child molestors.
              \_ I think you are underestimating the stupidity of the FBI.
        \_ This is nuts, BPD is out of hand here.
           \_ looks like this is the UCPD and the FBI, not BPD.  plus people
              have been harassing and leaving low yield pipe bombs in the yards
              of campus bio researchers.  I'm sure this is why they got raided.
              some dumbass used the free internet at the Long Haul to email out
              threatening letters.

8/24    what is iSCSI ?  thanks
        \_ http://en.wikipedia.org/wiki/Iscsi

8/6     funny: http://www.youtube.com/watch?v=OBbm8sV6IKk
        \_ no.

8/4     I went to Fry's for the first time in ages and was surprised to
        see a 1T Seagate HD on sale. I'm still using my 250G and it's
        only 60% full. What the heck do people put on their HD these
        days, porn? I can't imagine why anyone would want to buy a
        1T drive for personal use.
        \_ HDTV.
           \_ I agree. Recording HDTV shows takes up a lot of space.
              \_ You could fill a 1TB disk with 25 HD movies.
              \_ You could fill a 1TB disk with 40 HD movies.
        \_ "640K ought to be enough for anybody."  -- Bill G.
        \_ Dude, isn't MS Office 2008 like 600 GB or something?
        \_ Huge honkin' backups.
        \_ 6 years ago I wasn't sure why people needed a 10G hard drive.

7/9     SSD don't improve laptop battery life:
        http://www.tomshardware.com/reviews/ssd-hdd-battery,1955.html
        \_ Already debunked, please go away mr one line on link.
           \_ Really, I read a couple of articles that said SSD don't
              improve laptop battery life, especially on the MBA. If
              you have a url to a debunking article, I'd like to read
              it.
              \- i looked at some numbers on power draw for various
                 components [for large machine room planning] and i
                 was surprised at how little power disks pulled. say
                 compared to even memory. so i am not surprised about this.
                 BTW, the design of the new Berkeley-NERSC computer
                 bldg and machine room is kinda interesting ... like the
                 choice of axis w.r.t. to the sun, the calculation of how
                 many days in the bay area the outside temp is lower than
                 the targetted temp inside the machine room etc.
                 \_ Is this over-engineering? I tend to think so, because
                    the contents of the room will change radically over
                    relatively short periods of time.
                    \- this isnt a machine room with 1u and 3u machines
                       coming and going. This is for "a computer" with
                       a 8 figure price tag.
                       a 8 figure price tag. and besides, the DoE actually
                       does "energy research".
                       \_ That "computer" will be obsolete in less than a
                          decade.

7/8     You know I was hoping the government would procrastinate again
        and tell us that they're not going to force HD signals till
        2010 or something, but that hope is fading. Where can I get a
        HD box thingie, how do I get a $50 rebate, and how much are
        those things? ok thx.

5/29    Considering abandoning tape backups in favor of something like this:
http://www.thermaltakeusa.com/product/Storage/hdd_station/blacx/st0005u.asp
        Please convince me that I'm crazy.
        \_ why?
           \_ Why do I want you to convince me I'm crazy?
              \_ what are you trying to accomplish?  What is your context?
                 \_ Sorry: currently backing up 500GB per night on tape as
                    well as replicating to off-site (read: on other coast)
                    servers. Considering alternatives to continually upping
                    the number of tapes; thought nightly rsync might be a
                    better idea. A half-dozen SATA HDs and this might do
                    the trick.
                    \_ 500 GB PER NIGHT?!?  Are you actually generating that
                       much new data per day?
                       \_ Absolutely not, but we're assuming that we'll
                          have to use the backups in a vacuum so I'm
                          hesitant to rely on incrementals. This is why
                          the above solution is so tempting.
        \_ Depends on your needs. You gonna keep this thing spinning all
           the time? No offsite backups? Gonna run ZFS on it? How will
           you checksum your data? I think disk is becoming popular, but
           it's not a cure-all.
           \_ It sounds to me like he's planning on using it specifically
              for offsite backup.  You don't even have to bother taking the
              drive out of a typical server hotswap rails with this thing.

4/29    Any stock thoughts on ISLN (Isilon) and the stroage field in general?
        \_ Buy low, sell high.  You can never have enough storage.
        \_ I think that the basic idea of using commodity pc hardware to build
           scalable NAS replacements is sound and has the potential to
           supplant disk arrays in the same way that clusters of PCs replaced
           supercomputers, but I don't like Isilon's approach.  They don't
           scale well, and they're using too much proprietary hardware
           (single source, and not as cheap as real off-the-shelf PCs) to
           really deliver on the low cost promise.  As an investment, the
           greater concern is that Netapp is openly hostile to them (in a
           way that they are not toward other players in this space), and
           has demonstrated a willingness to sell at a loss to win deals
           away from them.
           \_ It actually boosts their credibility in my eyes if Netapp
              perceives them as a threat.
              \_ My impression that it's more personal animosity between the
                 management of the two companies than anything else.  I expect
                 the interesting products in this space to come from someone
                 else using real commodity hardware.
              \_ If Netapp is willing to give away filers to deny them
                 business, then they're a risky investment (even if they
                 are a credible technological competitor).

4/11    Countrywide 12 month CD is 4.25%. But some people tell me they're
        going out of business. Is it actually safe to deposit into
        Countrywide right now? Will FDIC pay me the full amount + interest
        should it ever goes out of business?
        \_ Are you really that concerned about 4.25%? I could see if it
           was 7% or something. How much more is that than the next best
           rate you can find? Anyway, the Countrywide web site says that
           is it is FDIC insured. FDIC will pay principal plus accrued
           interest up to a total of $100,000.
           \_ I'm putting down $20,000. Most rates are 3.00%, so by
              putting into Countrywide I'll gain an extra 1.25% interest
              rate or $250.00/year. That'll get me 2 good Chez Paniz
              meals. I'll take it.                              -op
              \_ Only if you eat by yourself and don't order wine, especially
                 after tax. How long will it take for you to recover your
                 money from Uncle Sam? Is $250 pretax really worth the risk?
              \_ Patelco CU ( http://patelco.org ) is having a new member special
                 right now.  6.785% for 12 month CD.  Existing members also
                 have a special rate of 4% 12 month CD for new monies.
                 \_ It says maximum amount of money you can put for 7%
                    is $1000.00. UH DUH.
                    \_ Ah, didn't see that the maximum is also $1000.  Sorry
                       about that.

3/12    Watch TV/movies from your desktop:
        http://www.hulu.com

2/25    Are there any DVD players that will remember the position of the
        last X movies played? I know most remembers the one currently
        being played, but what about previous ones? Thanks.
        \_ DVD Player in MacOS X remembers the last played position for
           all of my DVDs.
        \_ Yes. My old Sony player (second generation) remembers the
           positions of up to 5.

12/31  I am looking for a PC tech/guru as a goto-guy when problems happen,
       (hard disk failures, etc.) must be located in the bay area, rates
       negotiable.  -daveh@csua.berkeley.edu
        \_ Geek squad will also archive your porn for free!
        \_ Shouldn't everyone at a csua address be capable of this?
           \_ I'm guessing he's got a business and doesn't want to spend his
              own time fixing windows.

11/15   So if I have a process running in Linux and kill -9 isn't
        killing it, and killing its parent process didn't kill it,
        and now it's reading as though its parent process is 1,
        is there any way to kill it short of rebooting the machine?
        \_ Sure, but you dont want to be mucking around in kernel data
           strucures.  Most practical way is to reboot the machine.
        \_ Sure, but you dont want to be mucking around in kernel data
           strucures.  Most practical way is to reboot the machine.
           \_ Ah, good ol' Linux reliability. Just out of curiosity,
              what kind of process is this anyway?
              \_ This has nothing to do with "Linux reliability"; it has to
                 do with trying to kill a process that's blocked.  The most
                 common scenario is a disk wait; NFS server goes away or
                 physical I/O error on hard disk hangs up the process.  -tom
              \_ I believe some low range of numbers are reserved for
                 kernel processes only and will not let you kill it with
                 conventional kill method.
                  \_ no, you can kill any process with kill signals from the
                     root user.   However, kill is just a signal, and there
                     are paths (i.e. device wait) where processes are too
                     wedged to process the signal and die nicely.  -ERic
                     \_ The process in question is referencing a SCSI device.
                        Is there a way to kill the device wait? -op
                        \- the singal handler will not send a signal to a
                        \- the singal dispatcher will not send a signal to a
                           process in a disk wait. you could try while 1
                           kill -9 PID or force an umount of the fs ... but
                           that's all unpredictable. there are some super
                           hairy things you can do but they are beyond the
                           scope of the motd and are os dep.
                        \_ Plug the SCSI device back in.  -tom
                           \_ Never unplugged it. Turned it off and turned it
                              back on. No love.
                              \- it might be interesting to see what happens
                                 if you change run levels or unload the
                                 scsi kernel module [rmmod on AssOS].
                                 But as I said earlier, if you dont know
                                 why you are going in a disk wait ... it
                                 could be something obvious like removable
                                 media, netowkring going out etc ... or it
                                 could be unclear [failing disk?] ... that's
                                 what you should be trying to figure out.--psb
                              \_ Anyway, your fundamental problem is that
                                 your process is waiting on the SCSI device,
                                 and it won't go away untill the SCSI device
                                 unblocks.  Look at rescan-scsi-bus or
                                 something.  -tom
                  \- You should try to undnerstand why "kill -9" isnt
                     killing the process.
        \_ Short answer: No.  You're waiting on the drive.  If your proc is
           in locked in 'D' state, you're hosed.  Figure out what is wrong
           with your drive.
        \_ Followup: Rebooted the machine, and now all is well. Will explore
           other methods of downing this particular process in case this
           arises again. Thanks to all for suggestions and information. -op
           \_ Depending on exactly what is going on you may not be able to
              kill the proc.  At my first job I was 'tape back up guy' among
              other things.  We had first gen crappy tape drives that often
              just stopped responding to commands.  Usually power cycling the
              tape drive would clear the procs, but very rarely that wasn't
              good enough and a full reboot was required.  These were Suns.
              YMMV, but you may find there is no answer beyond 'reboot'.
              \- tape drive device drivers look more like disk device drivers
                 than tapes looked like disks. --psb
                 \_ Uhm, ok.  Yes.  Are you supporting what I said or
                    disagreeing in some way or ...?   --confused

11/12   What is diff between SAN and NAS?
         \_ NAS=Network Attached Storage.  Appliance which provides disk
             to servers/clients via file-based protocols (NFS, CIFS, iSCSI).
            SAN=Storage Area Netwoork.  Provides direct fiber connections from
             multiple servers to a single storage array.  -tom
            \_ Here is another way to explain It: They're very similar in that
               they are both technologies to have storage separate from the
               server. However with NAS, the storage protocols run over the
               server's normal network interfaces.  With SAN they go over
               special 'dedicated' storage-only connections, typically fiber.
               Furthermore, SAN uses a different storage-optimized protocol,
               whereas NAS typically operates over general network protocols
               such as IP. -ERic