lwt1973King of ThievesSyndicationRegistered Userregular
The controller is asking whether to take SQL for reporting or to go for Crystal Reports for reporting. He's had experience in prior versions of CR but that was a while ago. Anyone have an opinion on which direction I should tell him to take?
"He's sulking in his tent like Achilles! It's the Iliad?...from Homer?! READ A BOOK!!" -Handy
not a doctor, not a lawyer, examples I use may not be fully researched so don't take out of context plz, don't @ me
0
RandomHajileNot actually a SnatcherThe New KremlinRegistered Userregular
So we had a (Dell) server at a remote site that had open shares get hit by a CryptoLocker from a PC a couple weeks back. Long story short, I ended up reinstalling the OS on this machine. It's RAID6 so they could have a good amount of storage. When my coworker went to make a Ghost image of it (before the contractor came in to set up the software), at reboot, two drives "failed" and a third showed up as foreign. Since RAID6 can only tolerate two drive failures, it wouldn't boot. Now, in my 12 years here, I've never seen two drive failures at the same reboot (though I know it is possible). I've only had a couple times where a "good" drive went foreign, and it was when it was actually in the process of failing. They were able to get it back to booting by importing that foreign drive back as native.
Now, I'm not an expert on this stuff by any means, but that seems to me like something it flat out suspect with the PERC card. I mean, they're having us update the firmware on that PERC card, so that could be the problem, but oofa doofa. Basically 3 drive failures out of 8 on a single server at the same time makes me really worry about the other 50 servers we have...
It's the use of % and @ as basic syntax. Winds up looking like Perl with mystery operators everywhere, and doesn't otherwise bring anything better to the table then Python.
At this point I'm annoyed by anyone requiring dynamic linking. Go has become my go to scripting language because when you inevitably need to deploy it somewhere you don't need to fight the dependency monster.
So we had a (Dell) server at a remote site that had open shares get hit by a CryptoLocker from a PC a couple weeks back. Long story short, I ended up reinstalling the OS on this machine. It's RAID6 so they could have a good amount of storage. When my coworker went to make a Ghost image of it (before the contractor came in to set up the software), at reboot, two drives "failed" and a third showed up as foreign. Since RAID6 can only tolerate two drive failures, it wouldn't boot. Now, in my 12 years here, I've never seen two drive failures at the same reboot (though I know it is possible). I've only had a couple times where a "good" drive went foreign, and it was when it was actually in the process of failing. They were able to get it back to booting by importing that foreign drive back as native.
Now, I'm not an expert on this stuff by any means, but that seems to me like something it flat out suspect with the PERC card. I mean, they're having us update the firmware on that PERC card, so that could be the problem, but oofa doofa. Basically 3 drive failures out of 8 on a single server at the same time makes me really worry about the other 50 servers we have...
how big are the RAID arrays? RAID really starts to run into problem with huge disks because generally the fault tolerance for raid is very small, i.e. if it detects more than a couple bad blocks it'll mark the drive as failed. So if you're running a RAID5 and a drive fails, when you stick a new drive in and it starts to rebuild, if during that process it detects some bad blocks on another drive then poof, your RAID array goes tits up. Why this is more common on arrays with large spinning disks is that there are just physically significantly more blocks on every drive that can go bad, so the chances of running into a bad block on a drive are a lot higher just by sheer volume of them.
So we had a (Dell) server at a remote site that had open shares get hit by a CryptoLocker from a PC a couple weeks back. Long story short, I ended up reinstalling the OS on this machine. It's RAID6 so they could have a good amount of storage. When my coworker went to make a Ghost image of it (before the contractor came in to set up the software), at reboot, two drives "failed" and a third showed up as foreign. Since RAID6 can only tolerate two drive failures, it wouldn't boot. Now, in my 12 years here, I've never seen two drive failures at the same reboot (though I know it is possible). I've only had a couple times where a "good" drive went foreign, and it was when it was actually in the process of failing. They were able to get it back to booting by importing that foreign drive back as native.
Now, I'm not an expert on this stuff by any means, but that seems to me like something it flat out suspect with the PERC card. I mean, they're having us update the firmware on that PERC card, so that could be the problem, but oofa doofa. Basically 3 drive failures out of 8 on a single server at the same time makes me really worry about the other 50 servers we have...
how big are the RAID arrays? RAID really starts to run into problem with huge disks because generally the fault tolerance for raid is very small, i.e. if it detects more than a couple bad blocks it'll mark the drive as failed. So if you're running a RAID5 and a drive fails, when you stick a new drive in and it starts to rebuild, if during that process it detects some bad blocks on another drive then poof, your RAID array goes tits up. Why this is more common on arrays with large spinning disks is that there are just physically significantly more blocks on every drive that can go bad, so the chances of running into a bad block on a drive are a lot higher just by sheer volume of them.
Yeah, I mean, I get that, and they're huge arrays. But there was absolutely no indication until a random reboot, after I had rebooted it about 10 times last week diagnosing and reinstalling Windows. And to have a drive be marked foreign, when it had been fine, too. Just a really odd thing, and I'm very surprised they didn't replace the RAID card (yet).
The controller is asking whether to take SQL for reporting or to go for Crystal Reports for reporting. He's had experience in prior versions of CR but that was a while ago. Anyone have an opinion on which direction I should tell him to take?
I think learning actual SQL would be more useful. Isn't Crystal basically a fancy database viewer/query writer?
Better to know the underlying concepts that can be applied to any database software. If the dream job uses Tableau or Spotfire you don't want to know only Crystal.
life's a game that you're bound to lose / like using a hammer to pound in screws
fuck up once and you break your thumb / if you're happy at all then you're god damn dumb
that's right we're on a fucked up cruise / God is dead but at least we have booze
bad things happen, no one knows why / the sun burns out and everyone dies
+1
RandomHajileNot actually a SnatcherThe New KremlinRegistered Userregular
The controller is asking whether to take SQL for reporting or to go for Crystal Reports for reporting. He's had experience in prior versions of CR but that was a while ago. Anyone have an opinion on which direction I should tell him to take?
I think learning actual SQL would be more useful. Isn't Crystal basically a fancy database viewer/query writer?
Better to know the underlying concepts that can be applied to any database software. If the dream job uses Tableau or Spotfire you don't want to know only Crystal.
Further, knowing SQL (particularly joins/where clauses) will help you immensely in Crystal later on. Learning report writing in Crystal can give you bad habits on some of that stuff (like letting Crystal do the grouping rather than doing that in SQL where it is more efficient). Also, if he's had prior experience in Crystal, it really hasn't changed much, even since CR9 like 14 years ago. Just better ways to display and group stuff.
So we had a (Dell) server at a remote site that had open shares get hit by a CryptoLocker from a PC a couple weeks back. Long story short, I ended up reinstalling the OS on this machine. It's RAID6 so they could have a good amount of storage. When my coworker went to make a Ghost image of it (before the contractor came in to set up the software), at reboot, two drives "failed" and a third showed up as foreign. Since RAID6 can only tolerate two drive failures, it wouldn't boot. Now, in my 12 years here, I've never seen two drive failures at the same reboot (though I know it is possible). I've only had a couple times where a "good" drive went foreign, and it was when it was actually in the process of failing. They were able to get it back to booting by importing that foreign drive back as native.
Now, I'm not an expert on this stuff by any means, but that seems to me like something it flat out suspect with the PERC card. I mean, they're having us update the firmware on that PERC card, so that could be the problem, but oofa doofa. Basically 3 drive failures out of 8 on a single server at the same time makes me really worry about the other 50 servers we have...
how big are the RAID arrays? RAID really starts to run into problem with huge disks because generally the fault tolerance for raid is very small, i.e. if it detects more than a couple bad blocks it'll mark the drive as failed. So if you're running a RAID5 and a drive fails, when you stick a new drive in and it starts to rebuild, if during that process it detects some bad blocks on another drive then poof, your RAID array goes tits up. Why this is more common on arrays with large spinning disks is that there are just physically significantly more blocks on every drive that can go bad, so the chances of running into a bad block on a drive are a lot higher just by sheer volume of them.
Yeah, I mean, I get that, and they're huge arrays. But there was absolutely no indication until a random reboot, after I had rebooted it about 10 times last week diagnosing and reinstalling Windows. And to have a drive be marked foreign, when it had been fine, too. Just a really odd thing, and I'm very surprised they didn't replace the RAID card (yet).
Had two drives go on a RAID 5 in one night at a client once, and they hadn't been backing up their AD (this was their DC) so yeah, was a fun weekend. Dell was great tho, they replaced all drives, controller and cabling because it was pretty odd to have two go at once like that. Happened around 10pm, had my hardware all back in and running by midnight during a Canadian winter no less.
0
RandomHajileNot actually a SnatcherThe New KremlinRegistered Userregular
So we had a (Dell) server at a remote site that had open shares get hit by a CryptoLocker from a PC a couple weeks back. Long story short, I ended up reinstalling the OS on this machine. It's RAID6 so they could have a good amount of storage. When my coworker went to make a Ghost image of it (before the contractor came in to set up the software), at reboot, two drives "failed" and a third showed up as foreign. Since RAID6 can only tolerate two drive failures, it wouldn't boot. Now, in my 12 years here, I've never seen two drive failures at the same reboot (though I know it is possible). I've only had a couple times where a "good" drive went foreign, and it was when it was actually in the process of failing. They were able to get it back to booting by importing that foreign drive back as native.
Now, I'm not an expert on this stuff by any means, but that seems to me like something it flat out suspect with the PERC card. I mean, they're having us update the firmware on that PERC card, so that could be the problem, but oofa doofa. Basically 3 drive failures out of 8 on a single server at the same time makes me really worry about the other 50 servers we have...
how big are the RAID arrays? RAID really starts to run into problem with huge disks because generally the fault tolerance for raid is very small, i.e. if it detects more than a couple bad blocks it'll mark the drive as failed. So if you're running a RAID5 and a drive fails, when you stick a new drive in and it starts to rebuild, if during that process it detects some bad blocks on another drive then poof, your RAID array goes tits up. Why this is more common on arrays with large spinning disks is that there are just physically significantly more blocks on every drive that can go bad, so the chances of running into a bad block on a drive are a lot higher just by sheer volume of them.
Yeah, I mean, I get that, and they're huge arrays. But there was absolutely no indication until a random reboot, after I had rebooted it about 10 times last week diagnosing and reinstalling Windows. And to have a drive be marked foreign, when it had been fine, too. Just a really odd thing, and I'm very surprised they didn't replace the RAID card (yet).
Had two drives go on a RAID 5 in one night at a client once, and they hadn't been backing up their AD (this was their DC) so yeah, was a fun weekend. Dell was great tho, they replaced all drives, controller and cabling because it was pretty odd to have two go at once like that. Happened around 10pm, had my hardware all back in and running by midnight during a Canadian winter no less.
Yeah, above all, Dell has been great for our remote guy to work with on this. It means I don't have to drive 3.5hrs each way just to slap some drives in or whatever.
It's the use of % and @ as basic syntax. Winds up looking like Perl with mystery operators everywhere, and doesn't otherwise bring anything better to the table then Python.
At this point I'm annoyed by anyone requiring dynamic linking. Go has become my go to scripting language because when you inevitably need to deploy it somewhere you don't need to fight the dependency monster.
Ruby's blocks are better than python's functions, and ruby's object system is better integrated with the language. Those are the only things Ruby really does better. I get the hate for the sigils, but at the same time they're pretty useful.
On the subject of "fucking up" I once lost a CEO's month worth of calendar entries while trying to fix their palm pilot. Yeah, this was a while ago. Anyways, I did a push back and forth between the device and outlook using the pipe software (crazy) and all of a sudden August didn't have any appointments. Problem was I didn't know what was in there if anything, I just looked at the current month (July) and saw it had synced. The new HR person there called my boss and said they didn't want me to come onsite anymore. Not their contact person for our company, not the CEO - the brand new HR person who was a miserable piece of shit who wound up getting fired by the board years later. Anyways, my boss calls me and we meet to talk about what happened and after I run him through it he says, "Well, yeah, that makes sense. Ok, no worries." He calls the CEO, explains and says what happened and adds, "He's the only one who knows how everything there works. If you want, we can come over and just restore from backup." CEO brushes him off, "No worries, I can rebuild it from my paper dayplanner."
Fun fact: I was outside IT at the time, I am now the Network Admin at the company in question some 10 years later.
Another time I lost a business' entire store of documents and accounting because....I emptied the Recycle Bin. That was where they stored their files on their lone XP machine because the owner thought the documents would be safe in there, since no staff would ever go into it.
Anyways, last I checked NK2 files aren't a thing under 2013 and forward. I haven't dealt with one in years.
edit: warranty fun - once had a drive go in a IBM/Lenovo server (still IBM at the time I believe) and while RAID didn't collapse, but they hadn't sprung for hot spare soooooo yeah, exciting. Called IBM - "We don't have any of those drives available in Ontario." We had 4 hour service on that server, but I guess IBM's definition of 4 hour vs Dell's is "Well, we responded to your call in 4 hours and told you it would be 4 business days to get the drive to you." We were a third party doing their actual work and the hardware was provided by the company who had contracted us to handle it all. I convinced them to get a hot spare since the cost of the drive was a drop in the bucket, and on all servers going forward to spring for hot spares. The company that contracted us made better money selling IBM than Dell tho, so they went with more IBM the next hardware refresh.
Thanks for the feedback; the context was Im trying to decide which language to really get in to, I've got good familiarity with Ruby, but little with python. Python seems to be favored more for what I'm looking at so I'll probably go back to python
If you're more familiar with Ruby I would just go with that. It's not like either language is lacking in sysop tools. If you're working on Windows, neither have a default environment and both of them have tools for publishing standalone exes.
the empty recycle bin icon just looks so...sad...and...empty.
Right click desktop -> Personalize -> Change Desktop Icons -> Select Recycle Bin (empty) -> Click Change Icon -> Pick the Recycle Bin (full) icon -> There, now it never looks sad and empty.
If you're more familiar with Ruby I would just go with that. It's not like either language is lacking in sysop tools. If you're working on Windows, neither have a default environment and both of them have tools for publishing standalone exes.
It's really about market share, I'll need to be able to debug scripts so I'm trying to make sure where the most common is is where I'm most familiar.
Oh god. I woke up and had this terrible dream that I was back in 2000. Windows XP as far as the eyes can see.
I had this the other day! Only it wasn't a dream; it was a school corporation's entire business department and ohgodwhy. I found this out only after them throwing a fit about their server still running Server 2003. Well thank god the server that's only logged into by me and 1 other admin is now safe; no worries about those user-machines, I'm sure they'll be fiiiiiiiiiiine.
If your app needs an OCX installed, just give it to me and I'll make it happen.
If you want me to turn off all ActiveX safeguards because that's the only way you can figure out to make your app run and play nice with other software, your software is shit and your devs are shit and your company is shit.
Which to be fair, from all accounts this software widely regarded as the worst/cheapest in the industry.
The 10 million dollar census voting system set up by IBM for the Australian government's federal election this year was felled by not performing the simplest of IT fixes... And it's looking like the total cost is going to end up being triple the original price, who could possibly have seen that coming?!?
If your app needs an OCX installed, just give it to me and I'll make it happen.
If you want me to turn off all ActiveX safeguards because that's the only way you can figure out to make your app run and play nice with other software, your software is shit and your devs are shit and your company is shit.
Which to be fair, from all accounts this software widely regarded as the worst/cheapest in the industry.
Well to be fair, it's usually because the devs are hamstrung and need "to make it work with IE6 because half the companies still use it".
not a doctor, not a lawyer, examples I use may not be fully researched so don't take out of context plz, don't @ me
The controller is asking whether to take SQL for reporting or to go for Crystal Reports for reporting. He's had experience in prior versions of CR but that was a while ago. Anyone have an opinion on which direction I should tell him to take?
Do you need pixel perfect pdf reports? Do you need to schedule and distribute? Are you really just talking about producing excel dumps?
How much do you have to spend? What's your DB now? What are your core needs?
Crystal/BO is the bane of my existence currently. This space is FULL of options at all sorts of price points and capability levels. The biggest decision point deciders you'll hit is as follows:
1. What's the source DB? (If SQL Server, just do the MS PowerBI/Reporting stack, it's easy and integrates well, but lacks in scheduling/distribution and pixel perfect reports)
2. How full featured do you need your scheduling and distribution? (if not very, the field is much wider)
3. What security controls do you need? (if easy or simple, again the field is wide)
4. What Business Intelligence features are key to the groups you serve? (EG Data Exploration, visualizations, statistical suite integration etc.)
5. What are your future plans in this space?
Posts
something about ruby annoys me
Now, I'm not an expert on this stuff by any means, but that seems to me like something it flat out suspect with the PERC card. I mean, they're having us update the firmware on that PERC card, so that could be the problem, but oofa doofa. Basically 3 drive failures out of 8 on a single server at the same time makes me really worry about the other 50 servers we have...
This is a clickable link to my Steam Profile.
It's the use of % and @ as basic syntax. Winds up looking like Perl with mystery operators everywhere, and doesn't otherwise bring anything better to the table then Python.
At this point I'm annoyed by anyone requiring dynamic linking. Go has become my go to scripting language because when you inevitably need to deploy it somewhere you don't need to fight the dependency monster.
how big are the RAID arrays? RAID really starts to run into problem with huge disks because generally the fault tolerance for raid is very small, i.e. if it detects more than a couple bad blocks it'll mark the drive as failed. So if you're running a RAID5 and a drive fails, when you stick a new drive in and it starts to rebuild, if during that process it detects some bad blocks on another drive then poof, your RAID array goes tits up. Why this is more common on arrays with large spinning disks is that there are just physically significantly more blocks on every drive that can go bad, so the chances of running into a bad block on a drive are a lot higher just by sheer volume of them.
This is a clickable link to my Steam Profile.
I think learning actual SQL would be more useful. Isn't Crystal basically a fancy database viewer/query writer?
Better to know the underlying concepts that can be applied to any database software. If the dream job uses Tableau or Spotfire you don't want to know only Crystal.
fuck up once and you break your thumb / if you're happy at all then you're god damn dumb
that's right we're on a fucked up cruise / God is dead but at least we have booze
bad things happen, no one knows why / the sun burns out and everyone dies
This is a clickable link to my Steam Profile.
It isn't done because certain people aren't getting buy-in because they're abrasive and don't care what the users think.
Had two drives go on a RAID 5 in one night at a client once, and they hadn't been backing up their AD (this was their DC) so yeah, was a fun weekend. Dell was great tho, they replaced all drives, controller and cabling because it was pretty odd to have two go at once like that. Happened around 10pm, had my hardware all back in and running by midnight during a Canadian winter no less.
This is a clickable link to my Steam Profile.
Ruby's blocks are better than python's functions, and ruby's object system is better integrated with the language. Those are the only things Ruby really does better. I get the hate for the sigils, but at the same time they're pretty useful.
Fun fact: I was outside IT at the time, I am now the Network Admin at the company in question some 10 years later.
Another time I lost a business' entire store of documents and accounting because....I emptied the Recycle Bin. That was where they stored their files on their lone XP machine because the owner thought the documents would be safe in there, since no staff would ever go into it.
Anyways, last I checked NK2 files aren't a thing under 2013 and forward. I haven't dealt with one in years.
edit: warranty fun - once had a drive go in a IBM/Lenovo server (still IBM at the time I believe) and while RAID didn't collapse, but they hadn't sprung for hot spare soooooo yeah, exciting. Called IBM - "We don't have any of those drives available in Ontario." We had 4 hour service on that server, but I guess IBM's definition of 4 hour vs Dell's is "Well, we responded to your call in 4 hours and told you it would be 4 business days to get the drive to you." We were a third party doing their actual work and the hardware was provided by the company who had contracted us to handle it all. I convinced them to get a hot spare since the cost of the drive was a drop in the bucket, and on all servers going forward to spring for hot spares. The company that contracted us made better money selling IBM than Dell tho, so they went with more IBM the next hardware refresh.
Why don't you?
XBL:Phenyhelm - 3DS:Phenyhelm
You keep trying. I keep not being cut.
XBL:Phenyhelm - 3DS:Phenyhelm
the empty recycle bin icon just looks so...sad...and...empty.
It just wants love. Bowen keeps denying it love.
XBL:Phenyhelm - 3DS:Phenyhelm
Right click desktop -> Personalize -> Change Desktop Icons -> Select Recycle Bin (empty) -> Click Change Icon -> Pick the Recycle Bin (full) icon -> There, now it never looks sad and empty.
Ffffuuuuuuuuuuck
Offfffffffffffffffffff
It's really about market share, I'll need to be able to debug scripts so I'm trying to make sure where the most common is is where I'm most familiar.
yeah I don't think I'd agree to that one
I had this the other day! Only it wasn't a dream; it was a school corporation's entire business department and ohgodwhy. I found this out only after them throwing a fit about their server still running Server 2003. Well thank god the server that's only logged into by me and 1 other admin is now safe; no worries about those user-machines, I'm sure they'll be fiiiiiiiiiiine.
You mean they didn't also ask you to disable UAC and add the account to the administrators group? How security conscious of them.
And turn off the Windows firewall and disable anti-virus while installing.
If your app needs an OCX installed, just give it to me and I'll make it happen.
If you want me to turn off all ActiveX safeguards because that's the only way you can figure out to make your app run and play nice with other software, your software is shit and your devs are shit and your company is shit.
Which to be fair, from all accounts this software widely regarded as the worst/cheapest in the industry.
The 10 million dollar census voting system set up by IBM for the Australian government's federal election this year was felled by not performing the simplest of IT fixes... And it's looking like the total cost is going to end up being triple the original price, who could possibly have seen that coming?!?
Well to be fair, it's usually because the devs are hamstrung and need "to make it work with IE6 because half the companies still use it".
Do you need pixel perfect pdf reports? Do you need to schedule and distribute? Are you really just talking about producing excel dumps?
How much do you have to spend? What's your DB now? What are your core needs?
Crystal/BO is the bane of my existence currently. This space is FULL of options at all sorts of price points and capability levels. The biggest decision point deciders you'll hit is as follows:
1. What's the source DB? (If SQL Server, just do the MS PowerBI/Reporting stack, it's easy and integrates well, but lacks in scheduling/distribution and pixel perfect reports)
2. How full featured do you need your scheduling and distribution? (if not very, the field is much wider)
3. What security controls do you need? (if easy or simple, again the field is wide)
4. What Business Intelligence features are key to the groups you serve? (EG Data Exploration, visualizations, statistical suite integration etc.)
5. What are your future plans in this space?
He doesn't use keyboard shortcuts.
Fucking shoot me.
what?
there's not a whole lot of shortcuts you can use while writing batch scripts or while remoting in (keyboard input is really rudimentary usually)
He highlights them with the mouse, goes to edit, chooses copy, highlights where he wants to paste, goes to edit, chooses paste.