As was foretold, we've added advertisements to the forums! If you have questions, or if you encounter any bugs, please visit this thread: https://forums.penny-arcade.com/discussion/240191/forum-advertisement-faq-and-reports-thread/
We're funding a new Acquisitions Incorporated series on Kickstarter right now! Check it out at https://www.kickstarter.com/projects/pennyarcade/acquisitions-incorporated-the-series-2

The [sysadmins] Thread: Quick, hide your user friendly policies, Bowen is coming back!

17778808283100

Posts

  • CogCog What'd you expect? Registered User regular
    While we're on the subject of SQL, does anyone know of a simple and ideally free program to give a user a flat view and ability to edit the table data of dbf and mdf files? They don't need to be able to run queries, they just need to be able to view the contents of the table and make changes. And I refuse to give them the SSMS.

  • DietarySupplementDietarySupplement Still not approved by the FDA Dublin, OHRegistered User regular
    Cog wrote: »
    TL DR wrote: »
    So what do you do if you have a database, ostensibly created in and for SQL 2000, that will attach and complete a backup in SQL 2008, that will not attach in SQL 2000 citing 'error 602 could not find row in sysindexes'? Assume that you really, really need to be able to create a new SQL 2000 backup. The original server that it lived on is kaput.

    So let me try and understand: you had a SQL Server database, created in SQL Server 2000. You then detached it and attached the database to a 2008 server, but you can't attach the database to another SQL Server 2000 server?

    If the database was upgraded during the attaching process, you might be out of luck as databases can move forward but not backwards in versions. That's why it's best practice to create a backup of the database and then restore (but in your case if the server was borked, you might not have had a choice).

    Is the other SQL 2000 server running the same service pack and cumulative update?

    Failing that, you can script out each individual object in the database, create them, and then pump in the data via identity inserts. There's two ways to do that: either via the SSMS console or via powershell; I wrote some scripts that do that if you're interested (although you'd probably have to attach the database in 2008 on some server that supports powershell).

    DietarySupplement is pretty much on point here. There's no easy way out of this. Backups, Backups, Backups before you fuck with production data. You said the original server was kaput, where did you get this DB copy from?

    Incidentally, why can it only live in SQL 2000?

    An interesting question as well; although unless they were using features specifically deprecated in 2005 (although I can't think of any that would break if they weren't there). You could always restore the database to a 2008/R2 (but not 2012) server and put the database into compatibility mode 80 (SQL Server 2000). Note that compatibility mode is NOT the same as running it on an actual 2000 server, but it might be worth a shot for you.

    TL DR
  • CogCog What'd you expect? Registered User regular
    Cog wrote: »
    TL DR wrote: »
    So what do you do if you have a database, ostensibly created in and for SQL 2000, that will attach and complete a backup in SQL 2008, that will not attach in SQL 2000 citing 'error 602 could not find row in sysindexes'? Assume that you really, really need to be able to create a new SQL 2000 backup. The original server that it lived on is kaput.

    So let me try and understand: you had a SQL Server database, created in SQL Server 2000. You then detached it and attached the database to a 2008 server, but you can't attach the database to another SQL Server 2000 server?

    If the database was upgraded during the attaching process, you might be out of luck as databases can move forward but not backwards in versions. That's why it's best practice to create a backup of the database and then restore (but in your case if the server was borked, you might not have had a choice).

    Is the other SQL 2000 server running the same service pack and cumulative update?

    Failing that, you can script out each individual object in the database, create them, and then pump in the data via identity inserts. There's two ways to do that: either via the SSMS console or via powershell; I wrote some scripts that do that if you're interested (although you'd probably have to attach the database in 2008 on some server that supports powershell).

    DietarySupplement is pretty much on point here. There's no easy way out of this. Backups, Backups, Backups before you fuck with production data. You said the original server was kaput, where did you get this DB copy from?

    Incidentally, why can it only live in SQL 2000?

    An interesting question as well; although unless they were using features specifically deprecated in 2005 (although I can't think of any that would break if they weren't there). You could always restore the database to a 2008/R2 (but not 2012) server and put the database into compatibility mode 80 (SQL Server 2000). Note that compatibility mode is NOT the same as running it on an actual 2000 server, but it might be worth a shot for you.

    It should be in compatibility 80 by default if it came from a 2000 server.

  • Khade97Khade97 PE10, UKRegistered User regular
    Random issue I would like to put past you guys - nothing major but it is bugging me.

    We have a user that has a USB 2 hub, with 3 devices plugged into it (ethernet adapter, printer and sensor for KBaM). This is plugged into a USB 3 socket on their machine and all the attached devices work fine.

    I issued the user with a USB 3 portable hard drive so thought I may as well het them a USB 3 hub too. The HDD, ethernet and printer all work fine on the new hub, however the sensor is being odd. It senses the keyboard from its usual position (about 50cm away), however it will only sense the mouse if it is less than about 10cm away (it is normally closer to 70 cm away on the 2.0 hub). This happens even if the sensor is the only thing in the hub, and no matter which socket I use on the hub.

    I have put the user back to the USB 2 hub, but it is bugging me as to why the 3 was having issues - any thoughts?

  • CogCog What'd you expect? Registered User regular
    Khade97 wrote: »
    We have a user that has a USB 2 hub, with 3 devices plugged into it (ethernet adapter

    what

  • CogCog What'd you expect? Registered User regular
    Ok seriously though, have you looked to see if that model KBaM is USB 3 compatible?

    This is why I hate wireless KBaM. More hassle than it's worth, especially for desktop PC's. It's not like you're gonna pick up your goddamn tower and monitor and go sit on the couch or some shit. I will suffer a compact wireless mouse for a laptop, but if you have a port replicator you plug a fucking wired mouse and keyboard into that shit.

    Wireless is for travel.

    bowenLe_GoatDonovan Puppyfucker
  • LD50LD50 Registered User regular
    My guess is that the sensor isn't getting enough juice.

    CogboweniTunesIsEvil
  • CogCog What'd you expect? Registered User regular
    LD50 wrote: »
    My guess is that the sensor isn't getting enough juice.

    It's almost certainly this.

    Delmain
  • CogCog What'd you expect? Registered User regular
    Cog wrote: »
    Khade97 wrote: »
    We have a user that has a USB 2 hub, with 3 devices plugged into it (ethernet adapter

    what

    And you owe us all an explanation for this abomination.

  • DelmainDelmain Registered User regular
    It's almost certainly either a laptop with a dead ethernet jack or like, a MacBook Air with no ethernet because it's too thin for the jack.

    bowenAioua
  • CogCog What'd you expect? Registered User regular
    I'm so used to everything running under warranty and just getting swapped and shipped off for a new one. :P

  • Erin The RedErin The Red The Name's Erin! Woman, Podcaster, Dungeon Master, IT nerd, Parent, Trans. AMA Baton Rouge, LARegistered User regular
    A lot of newer things (slates, some tablets) have no onboard ethernet. Hence the need for a usb ethernet adapter.
    When I was tasked with finding a solution for getting windows 8 going for the whole company on the slates, it was a nightmare.
    Lenovo's ethernet adapter was usb 3, but would only work at usb 1 speeds when not actually IN windows. it made pxe booting to our mdt server take FOR. EVER.
    Then the higher-ups just recommended that we use usb sticks for all our deployments, which was sort of a dumb idea with how often we changed the image.

  • Khade97Khade97 PE10, UKRegistered User regular
    Cog wrote: »
    Cog wrote: »
    Khade97 wrote: »
    We have a user that has a USB 2 hub, with 3 devices plugged into it (ethernet adapter

    what

    And you owe us all an explanation for this abomination.

    One of the Surface tablets we just issued - all they have is a single USB port we are expanding out. If the sensor was getting power through USB2 with other things plugged in, why would it not get power through USB 3 with nothing else plugged in :neutral_face: They can just stick to USB2 for now I think

  • DelmainDelmain Registered User regular
    I'm going to go with "the USB2 hub was self-powered and the USB3 one is designed for external power"

    bowenLe_GoatDonovan Puppyfucker
  • TL DRTL DR Not at all confident in his reflexive opinions of thingsRegistered User regular
    TL DR wrote: »
    So what do you do if you have a database, ostensibly created in and for SQL 2000, that will attach and complete a backup in SQL 2008, that will not attach in SQL 2000 citing 'error 602 could not find row in sysindexes'? Assume that you really, really need to be able to create a new SQL 2000 backup. The original server that it lived on is kaput.

    So let me try and understand: you had a SQL Server database, created in SQL Server 2000. You then detached it and attached the database to a 2008 server, but you can't attach the database to another SQL Server 2000 server?

    If the database was upgraded during the attaching process, you might be out of luck as databases can move forward but not backwards in versions. That's why it's best practice to create a backup of the database and then restore (but in your case if the server was borked, you might not have had a choice).

    Is the other SQL 2000 server running the same service pack and cumulative update?

    Failing that, you can script out each individual object in the database, create them, and then pump in the data via identity inserts. There's two ways to do that: either via the SSMS console or via powershell; I wrote some scripts that do that if you're interested (although you'd probably have to attach the database in 2008 on some server that supports powershell).

    I have a file copy of the db files, and haven't done anything to modify the originals. This was taken from an NT backup that was run on the old server. They never performed a SQL backup, however.

    The app we're trying to coax back to life is so far out of support it's ridiculous. The only reason they're even talking to us is because the client is trying to move to the new app from the same vendor. That app is supposedly much more forgiving, so we're going to try to plug this db directly into it and see what happens.

    As I'm sure is obvious from my ineffectual flailing, I am not a DBA by any stretch. If this operation fails, I'll tell the client they're looking at $5k for RAID recovery or another 8-12 hours of my time while I learn up on SQL and PowerShell, possibly to be followed by $5k in RAID recovery anyway if that doesn't work.

  • CogCog What'd you expect? Registered User regular
    So you have a file copy of the db files that were essentially just attached to a 2008 server?

  • LD50LD50 Registered User regular
    Also, your clients should be slapped on the wrist for ever having used NTbackup. They are so fucking lucky that the backup was usable.

    Cog
  • DietarySupplementDietarySupplement Still not approved by the FDA Dublin, OHRegistered User regular
    edited March 2014
    Cog wrote: »
    So you have a file copy of the db files that were essentially just attached to a 2008 server?

    That's what I'm wondering: what version of SQL was running on the server that you restored from?

    And are you trying to attach the databases via SSMS or TSQL scripting alone?

    edit: If the former you may need to make sure you have the appropriate version of management studio installed; it can be touchy otherwise. For example, if it's a 2000 database I've seen instances where there are errors attaching the database to a 2008 or 2012 server, but if you use a 2000 or 2005 management studio it will work. I know, it's nutty, but that's how it goes...

    DietarySupplement on
  • CogCog What'd you expect? Registered User regular
    edited March 2014
    If that flat file copy was the only copy you had, and you attached it to the 2008 SQL server, you have probably made irreparable changes to it by simply attaching it to the 2008 server. You might have gotten away with making copies of the files and attaching those. It's a wildly non-ideal solution, but better than, again, dicking around with production data. Your best hope is probably reattaching to the 2008 server and scripting something to suck the raw data out and dump it into a new database on a new 2000 server.

    If I were you, I would tell these people to start asking themselves exactly how catastrophic a complete loss of data would be, and what steps they would need to take to begin to move forward anew in the eventuality that data recovery is really fully impossible. How irreplaceable and business critical is this data?

    And then, assuming NTBackup was not your decision, I would tell them that relying on it was foolish and risky beyond all reason and this sort of bullshit is why grownup goddamn companies use grownup goddamn backup solutions.

    Cog on
    uean
  • AbracadanielAbracadaniel Registered User regular
    IcyLiquid wrote: »
    is that an arc reactor next to a bottle of purel

    Lol. Rackspace sent me an Iron Man iPhone case for my birthday :p So yes thats an arc reactor :)

    Rackspace is good people, I've got a buddy who works for their office in SF. It's a very fancy office :D

  • TL DRTL DR Not at all confident in his reflexive opinions of thingsRegistered User regular
    I never touched my original copy - all attempts at fixes were done with separate copies of the data.

    None of this was our call. We were brought on to build them a new server that was to replace this server. They just failed to pull the trigger for 11 months and then stopped paying attention even to their NT backup. We're very lucky things are going as well as they are.

    I just sent their db to the app vendor, who seems confident that they'll be able to attach it to the appropriate version of SQL and make a backup. It being Friday, I am more than content to wash my hands of this.

  • CogCog What'd you expect? Registered User regular
    The best thing that I think could be done with this -- which I would personally rather lick call center keyboards than do -- would be to build a temporary server (preferably VM) for SQL 2000, attach the DB to that, take some REAL backups, and then perform the vendor approved upgrade path to whatever software replaced the old package, including appropriate DB upgrades along the way. As soon as the DB is in a 2008 ready state, move it to its new permanent home on a new 2008 R2 server and implement legit backups and maintenance plans.

    TL DR
  • CogCog What'd you expect? Registered User regular
    Also, I just had some pretty damn good cheesecake. That is how your bribe your IT department.

    I know which department's tickets are getting bumped to the top of the queue next week..

    AiouaErin The Red
  • Le_GoatLe_Goat Frechified Goat Person BostonRegistered User regular
    What's funny is that I just took my receptionist out for smoothies to let her know that I appreciate the extra help she gave me with box recalls from Iron Mountain last week during The Great Tape Exodus. I'm that type of person. You'd think people would maybe do the same....

    While I agree that being insensitive is an issue, so is being oversensitive.
  • CogCog What'd you expect? Registered User regular
    She thinks you were hitting on her.

  • LD50LD50 Registered User regular
    You think he wasn't.

    Delmain
  • Le_GoatLe_Goat Frechified Goat Person BostonRegistered User regular
    Wow, you seriously think that is what it was about? You honestly believe that I am that much of a disgrace? Allow me to be the first to tell you to go fuck yourself right up your arrogant ass. Honestly, fuck you.

    While I agree that being insensitive is an issue, so is being oversensitive.
  • GaslightGaslight Registered User regular
    Jesus dude I am pretty sure it was just supposed to be a joke.

    Delmain
  • LD50LD50 Registered User regular
    While it wasn't exactly my wittiest moment, it was meant to be tongue-in-cheek.

  • ueanuean Registered User regular
    So about that SQL upgrade, what if you.... Oh. Drama! I'll wait.

    Guys? Hay guys?
    PSN - sumowot
    TL DR
  • jeffinvajeffinva Koogler coming this summerRegistered User regular
    SCOM has all kinds of tests it performs with the Exchange management pack. It's a given that each environment is going to differ, in many cases in very nuanced ways, so these tests have to be tweaked (or outright disabled) in order to be effective. Unfortunately, I'm convinced certain inter-role communication mechanisms within an Exchange 2010 environment are just completely fucking unpredictable at times. Like the ECP latency tests. Under almost no load, with all conditions chopped down or excised, from CASs that are extremely powerful, the response times can vary by as much as 5000% within a 60 second sample. And then for many hours, or days, be completely normal.

    Nothing at all is to blame within any underlying or connecting infrastructure. Not VMware, the Big IP load balancers, SAN, nothing. Just fucking Exchange randomness. The answer I always get back from Microsoft is "*shrug* yea, that's the behavior we see too. We have reported it to the engineers - back in like 2003".

    I'm just going to start tallying up the hours spent verifying the environment following alarms alongside a count of how many times an alarm was valid, and submit it alongside a request to just disable about 90% of these monitors.

  • jaziekjaziek Bad at everything And mad about it.Registered User regular
    edited March 2014
    Oh god. Oh god.

    Just discovered that whoever set up the domain controllers (and one of the O&M servers, but that's not so much of an issue) did the RAID config wrong. The HDDs were built up on different hardware and then put into the server they were intended to be used in.

    There is RAID configuration on the drives, and the controller works, but then the mobo boots straight off on one of the drives instead of the controller.

    FUCK.

    Is this fixable? Take out the drive that has never been written to, import the config from the "main" drive onto the controller, controller thinks it's a degraded RAID, put the second one in and initialise it, add it to the now degraded VD? That would work right? If it doesn't we're well and truly fucked here.

    jaziek on
    Steam ||| SC2 - Jaziek.377 on EU & NA. ||| Twitch Stream
  • Mei HikariMei Hikari Registered User regular
    jaziek wrote: »
    Oh god. Oh god.

    Just discovered that whoever set up the domain controllers (and one of the O&M servers, but that's not so much of an issue) did the RAID config wrong. The HDDs were built up on different hardware and then put into the server they were intended to be used in.

    There is RAID configuration on the drives, and the controller works, but then the mobo boots straight off on one of the drives instead of the controller.

    FUCK.

    Is this fixable? Take out the drive that has never been written to, import the config from the "main" drive onto the controller, controller thinks it's a degraded RAID, put the second one in and initialise it, add it to the now degraded VD? That would work right? If it doesn't we're well and truly fucked here.

    Copy the entire RAID somewhere else, nuke the entire thing, redo it properly. Any other way will be like shooting yourself to get a bullet out.

    jeffinvaAiouaDonovan Puppyfuckerbowen
  • jeffinvajeffinva Koogler coming this summerRegistered User regular
    edited March 2014
    Edit: since I'm not sure what hardware you're using, changing my answer to: if this is not the only DC rebuild the array, build the server back up, and promote it again. No point in horsing around with that array. If it's your only DC back it up, verify the backup, rebuild, and perform a standard DC restore.

    jeffinva on
  • jaziekjaziek Bad at everything And mad about it.Registered User regular
    Now see, these seem like sensible solutions.

    Not something my manager is going to let me do.

    I don't get him. It seems like he knew what he was doing in the year 2000 and just hasn't bothered learning anything since then.

    Yesterday: Oh look, the move to sql server 2012 has broken some of the jobs on this server.

    What does this one do?

    BACKUP LOG <dbname> WITH TRUNCATE ONLY

    What does he tell me to do? Change it to

    BACKUP LOG <dbname> TO DISK='NUL:'

    DBCC SHRINKFILE (dbname_log,1)

    :rotate:

    Steam ||| SC2 - Jaziek.377 on EU & NA. ||| Twitch Stream
  • jeffinvajeffinva Koogler coming this summerRegistered User regular
    Yea, sometimes management types lose a lot of their technical prowess because they have to focus so much on esoteric managerial and "vision" crap.

    If you do only have one DC and your org is large enough to afford two (and really should have two), then you can mention this to him. Build up and promote a second DC, let the replication occur, leave the environment stable for a week, then rebuild that first DC on a proper array.

  • jaziekjaziek Bad at everything And mad about it.Registered User regular
    jeffinva wrote: »
    Yea, sometimes management types lose a lot of their technical prowess because they have to focus so much on esoteric managerial and "vision" crap.

    If you do only have one DC and your org is large enough to afford two (and really should have two), then you can mention this to him. Build up and promote a second DC, let the replication occur, leave the environment stable for a week, then rebuild that first DC on a proper array.

    Unfortunately, the system in question is part of a platform we sold to a customer and are contracted to maintain. Its a bad situation all around.

    I think technically the SLA doesn't cover proactive maintenance, and I was on there looking at an unrelated issue, so unless the DC dies, or one of their guys notices this problem, I'm more than likely gonna get told to keep it quiet.

    Steam ||| SC2 - Jaziek.377 on EU & NA. ||| Twitch Stream
  • Le_GoatLe_Goat Frechified Goat Person BostonRegistered User regular
    LD50 wrote: »
    While it wasn't exactly my wittiest moment, it was meant to be tongue-in-cheek.
    I work in a rather feminist environment. Every single day I have to deal with someone in the office making a comment about how I'm stupid because I have a different set of genitalia than they do. "Of course he fucked up. He's a man." Shit like that. I constantly have to watch what I say or do because a staff member could easily sway it in the direction of me being a sexist, chauvinistic bigot, and it pisses me off to no end. So when I read your comment, I quickly took offense due to my hypersensitivity towards the topic. There were over 100 more tactful ways for me to have responded, but I chose the more juvenile, lashing out approach. That being said, I apologize for the severity of my response.

    While I agree that being insensitive is an issue, so is being oversensitive.
  • SentretSentret Registered User regular
    Cog wrote: »
    While we're on the subject of SQL, does anyone know of a simple and ideally free program to give a user a flat view and ability to edit the table data of dbf and mdf files? They don't need to be able to run queries, they just need to be able to view the contents of the table and make changes. And I refuse to give them the SSMS.

    There's a reasonably good chance that you can do it with MS Access.

  • BigBadWolfBigBadWolf Grandma's HouseRegistered User regular
    Sentret wrote: »
    Cog wrote: »
    While we're on the subject of SQL, does anyone know of a simple and ideally free program to give a user a flat view and ability to edit the table data of dbf and mdf files? They don't need to be able to run queries, they just need to be able to view the contents of the table and make changes. And I refuse to give them the SSMS.

    There's a reasonably good chance that you can do it with MS Access.

    Is the database actually attached to a SQL server instance, or do just want to edit the data straight from the mbf files?
    If its not attached, a workaround might be be to install SQL server express (free), attach the database - then just link to the data from MS access.

    With appropriate permissions they can edit the data, but not touch the underlying structure (which i'm assuming is your concern with giving users SSMS).

This discussion has been closed.