• pedant (unregistered) in reply to Andy Goth
    Andy Goth:
    Quango:
    raid disk array
    I always chuckle inwardly whenever someone says "RAID array". Spelling it out even further ("RAID disk array") made me laugh out loud. Now I'm just waiting for someone to start talking about their redundant RAID. :^)
    "Redundant RAID" isn't complete nonsense. You might have a secondary RAID that acts as a fallback for a primary RAID.
  • (cs) in reply to D00fus
    D00fus:
    I suffer from RAS syndrome you insensitive clod!
    That was unnecessarily redundant repitition beyond what was needed.
  • Just Another WTF (unregistered) in reply to pedant
    pedant:
    Andy Goth:
    Quango:
    raid disk array
    I always chuckle inwardly whenever someone says "RAID array". Spelling it out even further ("RAID disk array") made me laugh out loud. Now I'm just waiting for someone to start talking about their redundant RAID. :^)
    "Redundant RAID" isn't complete nonsense. You might have a secondary RAID that acts as a fallback for a primary RAID.

    But then wouldn't the whole thing then just be another RAID level as in RAID 5+1 or 0+1. So in the end really all you have is RAID.

  • Sigh (unregistered) in reply to Nathan
    Nathan:
    Downfall:
    How is it that the WTF's alternate between "so poorly edited the reader can't figure out what caused the initial problem" and "so heavily edited that the original submitter shows up to complain and correct the record?"

    I agree with this. How hard is it to explain what happened without either: a. Fictionalizing it so much that it is unbelievable b. Having it make no sense.

    I miss the days when the DailyWTF was less "Story time with Alex" and more "Look at this bad code...make sure you're not dumb enough to do this!'

    At least the Error'd and CodeSOD sections still speak for themselves.

    I agree. I also miss that topcoder guy that used to yank everybody's chains. Fun times. Oh well, back to /.

  • (cs)

    RAIDs be damned.

    I fail to see what's so "whopping" about six and a half minutes. Christ, some of us know-nothings have had to fix nitwit DBA SQL that takes more than 1440 minutes to run the daily report. (Those at the back can put their shoes and socks back on now.)

    And yes, (a) The SQL statement is a WTF. Apart from anything else, the WHERE clause is wrong:

    WHERE RecordType = '02' AND RecordType
    -- previously defined, no doubt
    EQUALS ProviderNPI
    (b) Never rely on anybody called "Mort" for anything. (c) "Mort was so meticulous about it that he didn't even want people looking at the underlying table." Danger, Will Robinson! (d) You keep using that word. (Well, once...) I do not think it means what you think it means.

    Somewhere, up in the heavens, there is the sound of the other shoe dropping. Does a shoe that doesn't drop make a sound?

    Why, precisely, did this trivial piece of WTF SQL inconsequentia (Captcha!) cause the record actually to disappear?

  • (cs)

    How about a redundant RAID array of inexpensive discs?

  • Adrian Pavone (unregistered) in reply to Quango
    Quango:
    I'm a sucker for such a challenge.. Just ran this SQL on my workstation

    1000 selects by indexed NPI: 453ms 1000 selects by substring NPI: .. still waiting after 13 minutes

    My workstation: Core quad Q6600, 8GB ram, sql 2005 64bit, running Vista x64, raid disk array, NOT a slow system. What the heck kind of laptop do you have that does the first in 30ms?

    In regards to your substring one still running, with the stats he provided (on the 30ms for indexed machine), the substring version took over an hour.

  • Adrian Pavone (unregistered) in reply to Andy Goth
    Andy Goth:
    Quango:
    raid disk array
    I always chuckle inwardly whenever someone says "RAID array". Spelling it out even further ("RAID disk array") made me laugh out loud. Now I'm just waiting for someone to start talking about their redundant RAID. :^)

    I get the same feelings everytime someone says PIN number.

    However, with RAID, calling it a redundant RAID does server a purpose. While incorrect from the fact the first letter of RAID is "redundant" and so you are repeating the word, look at the various levels of RAID and what the term refers to these days. These days, RAID refers to an array of disks that have been combined to form a larger disk than any of the individual disks, OR it refers to mirroring (redundant disks). The various levels then build on these two concepts with more of one of the other, but at it's basis this is what RAID is these days.

    Now, with mirroring (RAID-1), that is a RAID in the true sense of the name (redundant disks). With striping (RAID-0), it is purely striping the data across all disks for speed and capacity, but if a single drive dies, everything is lost.

    So these days, calling something a redudant RAID serves the purpose of stating that you are using a RAID-1 (or one of it's various improvements, RAID-5, RAID-10, etc.)

    So while I agree with you from a pure theoretical level, the word RAID has taken on a meaning beyond what it's acronym stands for, and so from a practical level I have to disagree. This (acronyms not meaning what they stand for) happens all the time in computing, primarily due to the fact it is such a buzzword driven field.

  • Adrian Pavone (unregistered) in reply to Smyle
    Smyle:
    Andy Goth:
    I always chuckle inwardly whenever someone says "RAID array". Spelling it out even further ("RAID disk array") made me laugh out loud. Now I'm just waiting for someone to start talking about their redundant RAID. :^)

    I don't get it. Why are people making fun of my redundant RAID inexpensive disk array?

    Umm, are you people mental? RAID stands for Redundant Array of INDEPENDENT Disks. Where did you people get the idea it stands for inexpensive?

  • (cs) in reply to wee
    wee:
    greg:
    Maybe I'm just dumb, but what does the beginning of the story have to do with the end?

    What does a retarded table that is super slow to query have to do with the missing data?

    My thoughts as well. I've seen this before, and, yeah, it's a WTF. But what happened to the missing data?

    That my friend, is exactly what the WTF is. The punchline is, that there IS no punchline. Its a MetaWTF.

  • nilamo (unregistered) in reply to Andy Goth
    Andy Goth:
    Quango:
    raid disk array
    I always chuckle inwardly whenever someone says "RAID array". Spelling it out even further ("RAID disk array") made me laugh out loud. Now I'm just waiting for someone to start talking about their redundant RAID. :^)
    So, I've got this redundant RAID disk array. It's HAWT.
  • Me (unregistered) in reply to Adrian Pavone
    Adrian Pavone:
    Umm, are you people mental? RAID stands for Redundant Array of INDEPENDENT Disks. Where did you people get the idea it stands for inexpensive?

    Because that's what it stands for. It was changed by manufacturers, but the original term was inexpensive.

  • Department of Redundancy Dept. (unregistered) in reply to Johann
    Johann:
    Andy Goth:
    Quango:
    raid disk array
    I always chuckle inwardly whenever someone says "RAID array". Spelling it out even further ("RAID disk array") made me laugh out loud. Now I'm just waiting for someone to start talking about their redundant RAID. :^)

    So, anyway, I got these Redundant Array of Independent Disks Arrays going and...

    What kind of configuration do you have on your Redundant RAID Array of Inexpensive Disks configuration do you have?

  • Not as DULL as you (unregistered)

    I just found out why this comment is missing.

    It's because

  • Wireghoul (unregistered)

    So there was this comment that should have been made, but it disappeared. So I checked with the commenter dude who is really secretive about the comments and he said the comment had been made, but when I went back to check it the comment still wasn't there so I went looking for the comment.

    I looked under my desk, behind my monitor, scratched my head and sighed...

    FROGS!FROGS!FROGS!FROGS!FROGS!FROGS!FROGS!FROGS!FROGS!FROGS!

    Get it? I don't...

  • Andy L (unregistered)

    That reminds me of the time I was trying to figure out who stole my lunch. After I investigated for a while, I discovered that the refrigerator's compressor had a slow leak and was wasting electricity.

    Case Solved.

  • Andy L (unregistered) in reply to Adrian Pavone
    Adrian Pavone:
    Umm, are you people mental? RAID stands for Redundant Array of INDEPENDENT Disks. Where did you people get the idea it stands for inexpensive?

    Here is the paper where the acronym "RAID" was first invented. They probably didn't invent the concept, but this is where the modern idea of a RAID comes from.

    http://www.cs.cmu.edu/~garth/RAIDpaper/Patterson88.pdf

    What do you notice about the title of the paper?

    As noted above, drive manufacturers changed it because they didn't want "inexpensive" in the name of their high-priced disks, but that's a relatively recent development.

    -Andy

  • Real-modo (unregistered) in reply to The Linguist
    The Linguist:
    Anonymous:
    Quango:
    Vista x64, raid disk array, NOT a slow system.

    This supremely brilliant antithesis.

    http://en.wikipedia.org/wiki/Litotes

    Not a word I manage to work into conversation every day.

    Thanks! That was not uninformative.

  • Foo (unregistered)

    Isn't it obvious?

    The materialized view ate the punchline.

  • Dave (unregistered) in reply to MindChild
    MindChild:
    Well, it is to solve a little ambiguity. "I got a new RAID"... a new RAID what? A RAID card? A RAID cage? A RAID disk array? Understand yet?

    RAID flyspray, I believe.

  • Dave (unregistered) in reply to Robert
    Robert:
    Fibonacci:
    That's my NPI number!

    Well, too bad for you that it's not a valid NPI! :P 1123581322, on the other hand, would be just fine.

    That's the stupidest NPI I've ever heard in my life! The kind of thing an idiot would have on his luggage!

  • No one special (unregistered) in reply to Andy Goth

    It could be worse: An inexpensive array of redundant RAID disks.

  • citricsquid (unregistered) in reply to No one special
    No one special:
    It could be worse: An inexpensive array of redundant RAID disks.

    Even worse: An expensive array of redundant RAID disks.

  • (cs) in reply to citricsquid
    citricsquid:
    No one special:
    It could be worse: An inexpensive array of redundant RAID disks.
    Even worse: An expensive array of redundant RAID disks.
    Ah, an EMC SAN...
  • (cs)

    That story left me hanging. Rows are missing in the view, and the explanation is that the database scheme is extremely unscalable and very slow on large selects?

  • grapkulec (unregistered)
    1. W(here)TF is a punchline of this story?
    2. i had a time of my life reading all your comments. i'm very not ungrateful, although i had to explain my co-workers why i'm laughing like a madman
  • gurhall (unregistered) in reply to Me
    Me:
    Adrian Pavone:
    Umm, are you people mental? RAID stands for Redundant Array of INDEPENDENT Disks. Where did you people get the idea it stands for inexpensive?

    Because that's what it stands for. It was changed by marketing tards, but the original term was inexpensive.

    FTFY

  • gravis (unregistered) in reply to Code Dependent
    Code Dependent:
    Johann:
    So, anyway, I got these Redundant Array of Independent Disks Arrays going and...
    My RAID array is inexpensive.
    Bull. The disks might be "inexpensive" the complete RAID system is not.
  • Litote. (unregistered) in reply to gurhall
    gurhall:
    Me:
    Adrian Pavone:
    Umm, are you people mental? RAID stands for Redundant Array of INDEPENDENT Disks. Where did you people get the idea it stands for inexpensive?

    Because that's what it stands for. It was changed by marketing tards, but the original term was inexpensive.

    FTFY

    If ever there was a comment that wasn't undeserving of delicious blueness, it's this one. Or any other that uses the phrase 'marketing tards'.

  • (cs) in reply to Andy Goth
    Andy Goth:
    Quango:
    raid disk array
    I always chuckle inwardly whenever someone says "RAID array". Spelling it out even further ("RAID disk array") made me laugh out loud. Now I'm just waiting for someone to start talking about their redundant RAID. :^)

    Meet my redundant RAID inexpensive disk array!

  • The Scourge of Codd (unregistered)

    The sad thing is that such things are not done merely in the darkness by Mort, but are known and maintained by vendor-certified DBAs, and probably created by more of the same. Hell, sometimes they don't even do a Mort-style rip-n-replace, they update the records from SQL.

    And you can count on it that these folks are billed to the government at outrageous rates. You want to sound like the shrink in the epigram

    "Not 'Are you saved?' But in informal, Insistent tones, 'Brother, are you normal?'"

    --J.V. Cunningham (quoted from memory, so don't shoot.)

  • ath (unregistered)

    I read through all the comments looking for the punchline. At last, I found it.

  • Campbellm (unregistered) in reply to Andy Goth

    Sheesh.

    Redundancy is also important in grammar. Good grammar is all about making sure your reader or listener clearly understands what you are speaking about.

    So there's nothing wrong with saying "RAID Array" - it's clear and unambiguous.

    Make the grammar work first, optimise later.

  • Mr. Vocabulary (unregistered) in reply to Litote.
    Litote.:
    gurhall:
    Me:
    Adrian Pavone:
    Umm, are you people mental? RAID stands for Redundant Array of INDEPENDENT Disks. Where did you people get the idea it stands for inexpensive?

    Because that's what it stands for. It was changed by marketing tards, but the original term was inexpensive.

    FTFY

    If ever there was a comment that wasn't undeserving of delicious blueness, it's this one. Or any other that uses the phrase 'marketing tards'.

    Yeah, but you've overlooked the opportunity for a delicious, mellifluous, and brilliant portmanteau: "marketards".

    It just works. Unlike them.

  • wintermute (unregistered) in reply to Andy L
    Andy L:
    That reminds me of the time I was trying to figure out who stole my lunch. After I investigated for a while, I discovered that the refrigerator's compressor had a slow leak and was wasting electricity.

    Case Solved.

    Win.

  • Ozz (unregistered) in reply to Dave
    Dave:
    Robert:
    Fibonacci:
    That's my NPI number!

    Well, too bad for you that it's not a valid NPI! :P 1123581322, on the other hand, would be just fine.

    That's the stupidest NPI I've ever heard in my life! The kind of thing an idiot would have on his luggage!

    Dang - I'd better change the combination on my luggage...

  • (cs)

    I'm not a DBA, but I thought maybe the NPI was missing because-- oh, I dunno-- the query kept timing out?

  • (cs)

    This hits close to home. We have a DTS package that I wrote that checks every night to see if there's an updated NPI list on the source website. If there is, it downloads it and replaces the contents of our local NPI table with the new list.

    First, the NPI list is published in a non-standard CSV format where every field is quoted, regardless of the contents of the field. When there are special characters in the field like double-quotes (") or commas (,), they are NOT ESCAPED.

    We couldn't use the off-the-shelf CSV parsing capabilities via DTS, so I had to write some elaborate C# code to parse the blasted thing.

    On a more amusing note, empty fields are also ALWAYS quoted, even though there's clearly no reason. In a file with thousands of fields per record and millions of records per file, there are A BLOODY LOT OF QUOTES in that file. Needless to say, it has a pretty impressive compression ratio in the ZIP file that we download.

  • silent d (unregistered) in reply to Litote.

    'marketing tards' is a redundant phrase.

  • Gerald (unregistered) in reply to Anon
    Anon:
    Dan:
    greg:
    Maybe I'm just dumb, but what does the beginning of the story have to do with the end?

    What does a retarded table that is super slow to query have to do with the missing data?

    Agreed. Even if the table were fixed, I don't see how the missing number would magically appear, unless maybe it was missing due to column spacing. Did he ever find it? Inquiring minds want to know!

    I second these comments! Sure the database design is a huge WTF, but the story seems to be missing a punchline.

    I submitted the story and, unfortunately, it's more boring than many would think. This was a single solitary table sitting out there in a sea of other crap tables. It just happened to have a view sub-stringing the columns instead of a nice indexed table with proper column definitions. I was checking the process and view to see if I could find if the NPI wasn't listed because it was a process/table issue and took a look at the table and view definitions out of curiosity.

    The NPI was mising because the state kept sending us an incomplete file.

  • (cs) in reply to Gerald
    Gerald:
    The NPI was mising because the state kept sending us an incomplete file.

    Oh I get it, the WTF is that someone in state government screwed up! Ha ha ha!

    The only way that could be more unusual is if the agency in question was the DMV.

  • Magpie (unregistered) in reply to Gerald
    Gerald:

    I submitted the story and,

    ...snip...

    The NPI was mising because the state kept sending us an incomplete file.

    Well, that's no fun, now we don't have anything to make fun of. What are we going to do now?

    Oh, look, a tweet!

  • my wtf (unregistered) in reply to MindChild
    MindChild:
    Well, it is to solve a little ambiguity. "I got a new RAID"... a new RAID what? A RAID card? A RAID cage? A RAID disk array? Understand yet?
    Andy Goth:
    Quango:
    raid disk array
    I always chuckle inwardly whenever someone says "RAID array". Spelling it out even further ("RAID disk array") made me laugh out loud. Now I'm just waiting for someone to start talking about their redundant RAID. :^)

    Yo dawg I hear you like RAID.

  • Paladin (unregistered) in reply to Honeyman
    Honeyman:
    Even so doing something inefficiently in a slightly more efficient way is probably going to be slower than doing efficiently just once.
    A lookup over the indexed field does not care if the index is based on the field directly or on the function over the fields. Doing SELECT ... WHERE f1(f2(field)) = 'blaA' (if an index exists over the f1(f2(field)) ) is not "slightly more efficient" than doing SELECT ... WHERE field = 'blaB' (if an index exists over the field) - it is virtually the same.
    don't think MS SQL supports this.
    Even though I've never used MS SQL, I seriously doubt MS SQL doesn't support it - or some similar feature.

    I think that Paradox 3.0 supports this. I'm almost positive. I've never used it, and never looked at a manual or anything - it's just the feeling of certainty I have.

  • tinkerghost (unregistered) in reply to Fedaykin
    Fedaykin:
    Was the view created a materialized view or not? While the underlying implementation is pretty crappy, a materialized view would get rid of all the performance problems and would be a *really* good reason for the dev in question to insist that people only use the view.
    Um, only if your definition of 'good' includes "because I'm an utter incompetent". I have dumped XML straight to a database table - but only because they were error messages & the schema was randomly ignored by the other end. Other than specific instances like that, dumping range delimited strings into a database is verboten. String parsing like this is WORM - even if you build a materialized view, you're going to need to reparse & re-index everything as you're inserting it and that's going to increase your processing load every time you alter any data in the table.
  • Paladin (unregistered) in reply to tinkerghost
    tinkerghost:
    Fedaykin:
    Was the view created a materialized view or not? While the underlying implementation is pretty crappy, a materialized view would get rid of all the performance problems and would be a *really* good reason for the dev in question to insist that people only use the view.
    Um, only if your definition of 'good' includes "because I'm an utter incompetent". I have dumped XML straight to a database table - but only because they were error messages & the schema was randomly ignored by the other end. Other than specific instances like that, dumping range delimited strings into a database is verboten. String parsing like this is WORM - even if you build a materialized view, you're going to need to reparse & re-index everything as you're inserting it and that's going to increase your processing load every time you alter any data in the table.

    Boy you guys really have no idea of the power materialized views have in MS SQL 2000. They are awesome.

    Some fools will tell you they don't exist but ignore the haters. Indexed views and plain indexed derived tables with triggers are for chumps.

  • (cs) in reply to Luke
    Luke:
    I'd submit that RAID these days has lost its meaning, it's no longer a valid acronym, and so in some ways Redundant RAID does make some sense. RAID 0 provides no data redundancy, so you could argue either that a) it's not truly RAID and therefore should be called something else or that b) RAID really doesn't mean anything more than "a bunch of disks as a single logical disk" and therefore Redundant RAID is a perfectly valid way of referring to the old-school RAID we know and love. Also, Redundant RAID implies a tinge or sarcasm and therefore gets my vote.

    Actually if you look at the original RAID definition like a hardware geek, it makes perfect sense. RAID 1 doesn't stripe at all. RAID 2 stripes at the bit level. RAID 3 stripes at the byte level. RAID 4 stripes at the block level, and RAID 5 is RAID 4 with distributed parity. RAID 0 was defined later, and the numeric designation is no doubt at least partly tongue-in-cheek.

  • Picard (unregistered) in reply to Kensey
    Kensey:
    Luke:
    I'd submit that RAID these days has lost its meaning, it's no longer a valid acronym, and so in some ways Redundant RAID does make some sense. RAID 0 provides no data redundancy, so you could argue either that a) it's not truly RAID and therefore should be called something else or that b) RAID really doesn't mean anything more than "a bunch of disks as a single logical disk" and therefore Redundant RAID is a perfectly valid way of referring to the old-school RAID we know and love. Also, Redundant RAID implies a tinge or sarcasm and therefore gets my vote.

    Actually if you look at the original RAID definition like a hardware geek, it makes perfect sense. RAID 1 doesn't stripe at all. RAID 2 stripes at the bit level. RAID 3 stripes at the byte level. RAID 4 stripes at the block level, and RAID 5 is RAID 4 with distributed parity. RAID 0 was defined later, and the numeric designation is no doubt at least partly tongue-in-cheek.

    Good description. To expand upon this a bit, RAID 1 is mirroring. The best of the standard RAIDs, but the most expensive (double the array price, essentially). RAID 0 is also called JBOD...Just a Bunch of Disks, typically bonded together with an LVM, but not necessarily.

  • Picard (unregistered) in reply to Picard

    I correct myself. RAID 0 is striped, though brain-dead...

    According to Wiki,

    " RAID 0 "Striped set without parity" or "Striping". Provides improved performance and additional storage but no fault tolerance. Any disk failure destroys the array, which becomes more likely with more disks in the array. "

    Oddly enough, that's exactly how I define my logical vols on my Unix boxen. But...these vols reside on an EMC SAN, which does the mirroring (RAID 1).

  • (cs)

    Needs more conclusion.

Leave a comment on “The Secret DTS Package”

Log In or post as a guest

Replying to comment #:

« Return to Article