|
blackswordca posted:I forwarded the client email to the general manager with a complaint about being thrown under the bus. Well see what happens but im putting my money on "nothing" No, it won't be nothing. You see what you did there was entrapment again. Expect to be told as much.
|
# ? Mar 26, 2014 23:26 |
|
|
# ? May 21, 2024 13:59 |
|
8 hours of debugging came in... Two weeks ago some of our guys have been transfering a third party's SaaS that we bought to run 3 sites on an AWS instance that can only get 2 IP addresses but it needed three SSL sites. On installation we decided to just use 3 different ports to run the sites, and we decided on sequential ports starting on 443. At the time we also decided to just renew the SSL cert because it's about to run out. Fast forward to yesterday when they install the cert and create an image of the server, and when the server comes back up after imaging the deployment share stops working. Weiiiirddddd. There's absolutely nothing in the imaging process that should cause that, Event viewer shows a few Schannel errors without any useful information AT ALL, the errors that happen when using net use or explorer offer no help at all ("computer name does not exist" when using the IP address was one of the errors). What's worse is that the affected computer would happily access an SMB share on a different computer. Now, I am not a windows expert so this took me WAY longer to diagnose and fix, I spent a few hours on it before calling AWS, network captures showed nothing (the last packet is the affected computer simply TCP RSTing rudely after the other computer initializes an SMB handshake). Keep in mind we drilled into every networking aspect, including netsh resetting the ip stack, disabling tcp offloading, checking adapter orders, checking lanman configurations on the registry, checking stored credentials on the maching, checked GPO. Eventually we gave up for the night and I promised the AWS guys we'd send them a copy of the machine after rebuilding the servers today (yeah yeah I should have deployment automated on salt, I've been busy, sue me). Today while rebuilding the server one of the SSL sites didn't want to come up, the one assigned to port four hundred... and... fourty... five, suddenly everything loving clicked in my mind. The experts amongst you probably knew what was wrong on my second sentence. Jesus loving christ you'd think windows would complain a little louder that it couldn't reserve 0.0.0.0:445 for SMB. Not a goddamn useful peep on the event viewer or from any SMB client (you'd think they'd say "wtf that thing is not talking any SMB protocol I know" or anything other than "Name doesn't exist"). Much Macallan 12 for me tonight. (I don't deserve 18 for not checking ports on the IANA list) deimos fucked around with this message at 23:47 on Mar 26, 2014 |
# ? Mar 26, 2014 23:41 |
|
deimos posted:Two weeks ago some of our guys have been transfering a third party's SaaS that we bought to run 3 sites on an AWS instance that can only get 2 IP addresses but it needed three SSL sites. On installation we decided to just use 3 different ports to run the sites, and we decided on sequential ports starting on 443. At the time we also decided to just renew the SSL cert because it's about to run out. I knew where this was going. Don't use ports under 1024 for things like this.
|
# ? Mar 26, 2014 23:51 |
|
evol262 posted:I knew where this was going. Don't use ports under 1024 for things like this. Yeah I usually don't, temporary lapse in judgement.
|
# ? Mar 26, 2014 23:58 |
|
A ticket came in... Email signature is not working for a client. Go through the basic troubleshooting, ask them to check if the signature is enabled in Outlook--it is. Then he asks me to check if any other tickets came in from anyone else, as he said he has not seen any signatures attached to emails for several days. Lo and behold, there's one other ticket in there. From the same user. Sent in 15 minutes ago. So I close the ticket as a duplicate, and state that all inquires be made in the new ticket, which has the Exchange provider attached to it. No more than 5 minutes later, I get a harshly-worded reply essentially saying "ISSUE NOT RESOLVED FIX THIS NOW!!!!!". Can I be any more clear than "we don't support this, and you have another ticket open on this already in our system and with the vendor"?
|
# ? Mar 27, 2014 00:45 |
|
blackswordca posted:I thought these people were fiction Oatmeal Raisin loving rocks.
|
# ? Mar 27, 2014 00:58 |
|
A post-migration ticket came in... Company had been migrated to a new version of a web app we offer effective that morning, users are unable to login. After some investigation, I found that the Project Manager had created their accounts and sent their temporary login information in February. Almost none of the users still had or knew where to find their login info, and for those that could it was not working. This was because temporary passwords expire after 30 days and it was more than 30 days from account creation that the site went live. We reset all their passwords, but about 10% of the users couldn't login with their new information. Come to find out that if the randomly generated temporary password contains a $ character, it will not be saved OR sent via e-mail correctly. This issue was encountered during testing and dismissed as user error, and no one ever noticed the passwords being saved and sent were obviously the wrong length. I'm just ignoring the fact that our company's security practices are frozen in the year 2001 in addition to not actually working.
|
# ? Mar 27, 2014 01:37 |
|
KoRMaK posted:Thank them for contacting you and cordially suggest that they submit an email to the new ticket system in order for it to be tracked. We did a phased approach. First it was, "in the future, please open a ticket." Then it was "Before I look at this, is there a reason you weren't able to open a ticket?" Only after a few months did we officially refuse to do work without a ticket. I was managing the UNIX sysadmin team at the time and the reaction was interesting - some of the guys were sick of some of the problem children in our organization and were delighted to have any excuse to tell them off, while others hated the idea that the couldn't fix problems that people brought to their attention. I had to make it very clear that anyone caught even touching a server without either a trouble ticket or a change ticket would be on my poo poo list, and I contacted the management for some of the worse offenders to make sure they understood that the entire IT organization had switched to using the ticket system, and that if their staff did not follow suit, their staff would be blamed for the delay in resolving the problem, because my clock did not start until the ticket was assigned to my team. I acted sad about it but honestly, I love cracking down on people who can't follow simple instructions.
|
# ? Mar 27, 2014 01:57 |
|
Roargasm posted:I would continue to take emails and IMs and have the IT staff log those calls themselves honestly As someone who used to do IT for a larger public high school, I absolutely refused to create my own tickets. If you can't be arsed to take the extra 60 seconds to pull up a website and log a ticket, I'm not going to do it for you. This mindset may have been affected by a lot of "I'm too busy to properly describe or reproduce my problem, just fix it" calls.
|
# ? Mar 27, 2014 05:00 |
|
Before we merged with another organization we had a 'helpdesk' line that any of our techs could answer and just kind of took jobs round-robin style. After merging we started directing everyone to a centralized helpdesk. It's been a year and a half since anyone has answered that line and we STILL get people calling it probably once a day. I like to imagine there's some guy who has been calling once a week and thinks he just gets unlucky every time and is still holding out hope that eventually he'll get through.
|
# ? Mar 27, 2014 07:27 |
|
Priss In Plate posted:A ticket came in... Can you not link or merge tickets?
|
# ? Mar 27, 2014 08:52 |
|
Fellatio del Toro posted:Before we merged with another organization we had a 'helpdesk' line that any of our techs could answer and just kind of took jobs round-robin style. We currently do our ticketing 95% through the phone and 5% through email and it works pretty well. Most of the time, issues can be immediately resolved over the phone so it seems pointless to have someone submit a ticket and then track them down to resolve what could have been a 2 minute fix. We are consultants for a smaller part (1200 users) of a very large (120k user objects in AD) org and the reason we get kept on is that we attempt to have every issue that isn't a project resolved within 6 business hours. If someone in one of the buildings next door submits a ticket for something simple like getting a piece of software installed that has a legitimate business need, they often have to wait days. I understand that our situation might be a bit different than most. We have 2 helpdesk dudes, two application/server guys, a DBA/developer, a helpdesk manager/acquisition guy, and a director. It never really feels overwhelming with the support ratio to be honest.
|
# ? Mar 27, 2014 14:23 |
We're a small shop, like 60-80 people in the entire company, majority in the office where I'm at. My boss was pushing them to do tickets since I started in November, and he sent out a nice reiterative email because our users seem to have taken "please submit a ticket" as "Please don't tell IT about anything until you're really frustrated by it" quote:Do's It's good carte blanche to fall back on. Of course, people come up to me and co-worker with stuff, or call us. It's good to be able to say "Okay, got the info - now what I need for you to do is submit a ticket more or less detailing what you just told me and we can get started on it" and then it goes through proper channels. Help Desk guy works the issue and escalates to me/checks with me as needed, I can immediately claim more critical stuff, we justify our existence, bada bing.
|
|
# ? Mar 27, 2014 16:03 |
|
So while I am talking with vendors to ensure we have enterprise monitoring purchased... a loving data store all of a sudden is showing over provisioned and fills up. I stop everything to check on it and for some loving reason, the disk files are duplicated in the store. What could cause this you ask? Backup-exec
|
# ? Mar 27, 2014 17:20 |
Spent all morning trying to figure out why a client's DateTimes were getting the month and day transposed, guess where their servers are [not] hosted? gently caress you America, why you gotta do everything rear end backwards.
|
|
# ? Mar 27, 2014 17:28 |
|
dennyk posted:Not entirely accurate; management has decided that they don't want to *pay* for 24x7 coverage. I'm sure it's still needed, though, which means you and the other remaining salaried exempt employees will be providing it (in addition to your normal daytime work, of course). This would be entertaining and equally impossible. I'd have so much leverage with what I do for a significant raise and/or quitting that its not even funny in addition to not even having been trained for the biggest impact part of what they do. That would be the "we're going to talk about how a 20k raise is going to define my continued employment." I wouldn't count it out 100% but I wouldn't expect that to happen.
|
# ? Mar 27, 2014 17:38 |
|
demonachizer posted:We currently do our ticketing 95% through the phone and 5% through email and it works pretty well. Most of the time, issues can be immediately resolved over the phone so it seems pointless to have someone submit a ticket and then track them down to resolve what could have been a 2 minute fix. You're getting 5% of the credit you're due. It might work well now, but if it ramps up and you need help, it might be awkward to ask for more staff when you only do 1/20 the industry standard workload according to your metrics. Tickets will also help you analyze large scale events better than "hey, anyone else seem to be getting a lot of calls about FooMunger throwing errors?" Always make tickets, even if it's you making a ticket after the fact just to document what happened.
|
# ? Mar 27, 2014 17:42 |
Manslaughter posted:Spent all morning trying to figure out why a client's DateTimes were getting the month and day transposed, guess where their servers are [not] hosted? gently caress you America, why you gotta do everything rear end backwards. When I took French in high school it was explained to me how the rest of the world (or at least Europe) writes dates and why they do it that way. I now have to constantly stop myself writing dates that way because it just makes so much more sense.
|
|
# ? Mar 27, 2014 17:44 |
|
Manslaughter posted:Spent all morning trying to figure out why a client's DateTimes were getting the month and day transposed, guess where their servers are [not] hosted? gently caress you America, why you gotta do everything rear end backwards. lovely is actually a huge compliment. It was so bad it couldn't even calculate GST correctly.
|
# ? Mar 27, 2014 18:09 |
|
President Ark posted:When I took French in high school it was explained to me how the rest of the world (or at least Europe) writes dates and why they do it that way. I now have to constantly stop myself writing dates that way because it just makes so much more sense.
|
# ? Mar 27, 2014 18:11 |
Collateral Damage posted:I have an deep hatred for anyone who doesn't use ISO8601 for date fields in internal data formats. I don't care how you present it to the user, but if your systems are sending data files to my systems with m/d/y or some other rear end-backwards date format in file names or data fields meant to be machine-readable I'll hunt you down and gently caress you with a rusty shovel. Oh, I'm not talking about datasystems, just handwriting or manual entry on forms or whatever.
|
|
# ? Mar 27, 2014 18:12 |
|
Manslaughter posted:Spent all morning trying to figure out why a client's DateTimes were getting the month and day transposed, guess where their servers are [not] hosted? gently caress you America, why you gotta do everything rear end backwards. As an American who works with software written by an Australian company, I say you all have it backward. Also, I'd much prefer if we switched to using YYYYMMDD for everything, internally and externally.
|
# ? Mar 27, 2014 18:28 |
|
Dragyn posted:As an American who works with software written by an Australian company, I say you all have it backward. As an American who works with software at all, if you're not storing your datetimes in YYYY-MM-DD HH:MM:SS, then you're all nuts
|
# ? Mar 27, 2014 18:30 |
|
I am the victor of Tablet Wars 2014! Long story short, sales department wants tablets for all personnel. My supervisor, the Chief Engineer, is pushing iPads and I, the one-man IT department, am pushing Surfaces. After a month-long war of attrition, the General Manager contacted me and told me to purchase, set up, and train the sales department to use fifteen Surfaces. I really thought this would piss off my supervisor, but he was actually so impressed with my Surface presentation that he's buying one for himself. Taco Bell has breakfast and now someone listened to a thing I said. This has been a pretty good day
|
# ? Mar 27, 2014 18:38 |
|
See if you can budget the extra $200 per Surface for an additional year of warranty, which also includes accidental damage protection. You can go here to buy that and register your Surfaces. Sales people tend to not take care of their poo poo, so this might be a worthwhile investment. https://myservice.surface.com/en-US/pages/welcome.aspx
|
# ? Mar 27, 2014 18:44 |
|
larchesdanrew posted:I am the victor of Tablet Wars 2014! Long story short, sales department wants tablets for all personnel. My supervisor, the Chief Engineer, is pushing iPads and I, the one-man IT department, am pushing Surfaces. After a month-long war of attrition, the General Manager contacted me and told me to purchase, set up, and train the sales department to use fifteen Surfaces. The winning move was telling them no.
|
# ? Mar 27, 2014 18:45 |
|
GreenNight posted:See if you can budget the extra $200 per Surface for an additional year of warranty, which also includes accidental damage protection. You can go here to buy that and register your Surfaces. Sales people tend to not take care of their poo poo, so this might be a worthwhile investment. This is the best part of the story, but since the sales department was so adamant about getting tablets, there's now a "you break it, you bought it" policy. They take care of their poo poo or it's coming out of their paycheck. If they don't agree, they don't get a tablet.
|
# ? Mar 27, 2014 18:45 |
|
larchesdanrew posted:This is the best part of the story, but since the sales department was so adamant about getting tablets, there's now a "you break it, you bought it" policy. They take care of their poo poo or it's coming out of their paycheck. If they don't agree, they don't get a tablet. Yeah but what if the Surface dies at month 14? Just a hardware failure, no damage. We've had some Surface Pro's that happened to. No dead Surface Pro 2's yet, but those are still new.
|
# ? Mar 27, 2014 18:48 |
|
larchesdanrew posted:This is the best part of the story, but since the sales department was so adamant about getting tablets, there's now a "you break it, you bought it" policy. They take care of their poo poo or it's coming out of their paycheck. If they don't agree, they don't get a tablet. "Oh, the client tripped over the connector, it's not my fault!", words said verbatim to me to convince me not to bill for a CATO laptop with TIRE MARKS on it
|
# ? Mar 27, 2014 18:48 |
|
Collateral Damage posted:I have an deep hatred for anyone who doesn't use ISO8601 for date fields in internal data formats. I don't care how you present it to the user, but if your systems are sending data files to my systems with m/d/y or some other rear end-backwards date format in file names or data fields meant to be machine-readable I'll hunt you down and gently caress you with a rusty shovel. You'd be surprised the crazy poo poo some platforms use for dates internally. ANSI M was the fever dream of some guy designing database platforms for hospitals, so he decided that the date 0 should be the day before the birthday of the oldest currently living American and that dates should be the number of seconds since that time. Pretty much any date coming in or out of software implemented on an M-based platform as to do hilarious work to figure out what format they're receiving or need to send in and translate. This is made more hilarious by the fact that several major banks also run software that uses M (why????), so they have to deal with the weird epoch and the fact that all numerical values are stored as doubles and there's no built in way to truncate them precisely, which makes dealing with currency amounts weird. I'm sure I've posted about this before, but it never stops being hilarious seeing the workarounds programmers have to use to get things that are completely standard in most modern database platforms.
|
# ? Mar 27, 2014 18:58 |
|
LeftistMuslimObama posted:You'd be surprised the crazy poo poo some platforms use for dates internally. ANSI M was the fever dream of some guy designing database platforms for hospitals, so he decided that the date 0 should be the day before the birthday of the oldest currently living American and that dates should be the number of seconds since that time. Pretty much any date coming in or out of software implemented on an M-based platform as to do hilarious work to figure out what format they're receiving or need to send in and translate. what the gently caress
|
# ? Mar 27, 2014 19:42 |
|
LeftistMuslimObama posted:You'd be surprised the crazy poo poo some platforms use for dates internally. ANSI M was the fever dream of some guy designing database platforms for hospitals, so he decided that the date 0 should be the day before the birthday of the oldest currently living American and that dates should be the number of seconds since that time. Pretty much any date coming in or out of software implemented on an M-based platform as to do hilarious work to figure out what format they're receiving or need to send in and translate. I have to translate all the dates in my system from their weirdass epoch date using excel... =TEXT((K2-21548),"mm/dd/yyyy") I think it's day since 12/31/1889 or something like that.
|
# ? Mar 27, 2014 20:17 |
|
Alctel posted:what the gently caress Actually it's a bit more than that, he found the date of birth of what was currently known to be the oldest living American, and then added a few years of buffer time beforehand in case someone turned out to have been born earlier than that. The epoch marker was decided in 1969, when the current oldest person in America was 121 years old i.e. born sometime in 1848; just to be safe the start date was pushed back to the last day of December 1840. You know, in case a 128 year old showed up at the hospital in 1969 shortly before dying of extreme age.
|
# ? Mar 27, 2014 21:23 |
|
Install Windows posted:Actually it's a bit more than that, he found the date of birth of what was currently known to be the oldest living American, and then added a few years of buffer time beforehand in case someone turned out to have been born earlier than that. The epoch marker was decided in 1969, when the current oldest person in America was 121 years old i.e. born sometime in 1848; just to be safe the start date was pushed back to the last day of December 1840. How is that even a helpful epoch marker in any way to begin with?
|
# ? Mar 27, 2014 21:27 |
|
Why not? There are currently two living grandchildren of John Tyler, and his presidency started in 1841. I just hope noone needs to refer to his birthdate in 1790.
|
# ? Mar 27, 2014 21:31 |
|
Alctel posted:what the gently caress MUMPS stories are always the funniest stories ever.
|
# ? Mar 27, 2014 21:33 |
|
Inspector_666 posted:How is that even a helpful epoch marker in any way to begin with? http://en.wikipedia.org/wiki/Epoch_%28reference_date%29#Notable_epoch_dates_in_computing Eh, less arbitrary than a lot of them, it's just that Unix epoch is much more common. There are plenty of more serious issues with MUMPS.
|
# ? Mar 27, 2014 21:38 |
|
scroogle nmaps posted:http://en.wikipedia.org/wiki/Epoch_%28reference_date%29#Notable_epoch_dates_in_computing I'd say it's pretty drat arbitrary given that he pretty much just took a wild guess on a date that is itself probably mostly apocryphal.
|
# ? Mar 27, 2014 21:40 |
|
Paladine_PSoT posted:Why not? There are currently two living grandchildren of John Tyler, and his presidency started in 1841. I just hope noone needs to refer to his birthdate in 1790. Just use negative time duh
|
# ? Mar 27, 2014 21:43 |
|
|
# ? May 21, 2024 13:59 |
|
Inspector_666 posted:How is that even a helpful epoch marker in any way to begin with? MUMPS was designed for use in a hospital environment, there was currently at least one guy still alive who was over 120 years old, so the start date should be set for before he was born because negative time values are icky or something and the designer wanted the oldest guy alive to be able to be admitted to the hospital's systems with his true age. Then the designer decided "why not add on a few more years".
|
# ? Mar 27, 2014 21:52 |