|
NevergirlsOFFICIAL posted:Look at JIRA I guess if you do any in house development. yes it is one of the few good small shop memories I have. Other small shops stuff includes (free or very cheap license of) Zendesk which I disliked because there was very little customization or exportability. You have to do things their way or pick another product. Used nagios for monitoring, but it's way too fiddly compared to PRTG for simple stuff. Microtik routers are a good choice provided you have someone around who knows how to use them but that's hardly ticketing or monitoring related.
|
#
¿
Apr 19, 2017 00:07
|
|
|
|
| # ¿ Apr 23, 2024 17:09 |
|
|
sneakyfrog posted:loving lol I can't tell if you're kidding because i know dont work with large (amounts of) batteries but in case you're not  The Li ions are pretty reactive but i thought that is fixed basically ~100% by the time the initial burning up completes and its all settled down into some used up oxidated state. Many months up to a year of cleanup makes it sound like it's suspended in a crystal lattice of anthrax and asbestos. Do you know what causes such a mess?
|
|
#
¿
May 8, 2017 21:31
|
|
|
That sounds familiar, ha. I was mainly curious about the hazardous bits since i figured there have to be at least some relatively bad ones for the cleanup to be such a hassle. I guess i need a vacation because assuming that any given IT/business process has a specific purpose or mechanisms to prevent terrible trainwrecks gets proven wrong spectacularly at some point. Actually i just got out of one such mess a week ago, a bit bigger than small <100 user IT but appropriate. I completed an 8 month journey through red tape, proposals and change boards, sanity checks and whatnot because 'I' 'wanted' 4x1 TB of empty unformatted disks, each attached as a mountpoint to a particular server in an Exchange cluster hosting 16 identical ones already. It's not like working 'with' agile/scrum/itil/voodoo/... Is new to me either. I know gets work done and who i should avoid. Unsuccessful this time,I think two external companies (not counting my own) were involved in that one, + the client of course. Maybe they called Microsoft at some point as well to give their OK on attaching a new mount point from a storage array doing not very much to a bunch of servers that happen to be low on database space. Don't remember exactly any more but probably this simple request slipped right out of my control because i shot the storage guys an email right before my vacation mentioning extra disk space in the close future so I wouldn't catch them by surprise. Rookie mistake, the wrong guy read it, misunderstood, sent it to the wrong manager who did the same and there you are. The client's doing good so the next cluster will probably need extra disks too very soon. Good lord
|
|
#
¿
May 8, 2017 22:53
|
|
|
I was at a small business today. Read their news feed on the intranet because why not. A couple weeks ago someone wrote 'hey guyz I made a powershell script repository so we don't all reinvent the wheel twice a week so if you've got anything nice throw it in there.' this is a decent idea, even if you're just moving stackoverflow answers to a network location. But where would be the best place to put them? That it's powershell might be a hint. Think hard. It's the age of disruption, baby. The answer is so disruptive that it's probably best to stop reading if you get angry really easily he put it on Yammer  of course it's empty
|
|
#
¿
May 24, 2017 00:42
|
|
|
Yes, what you're doing is going to cost some money. Depending on what hardware they have now, maybe you can get a decent setup going without scaring your boss/client shitless. 10 people isn't a lot, however. Setting up small domains + hardware isn't really my expertise but here are some things I've run across in this general area. Might be a lot of dinosaur corporate IT ideas in here though Office365 is the platform that includes email via Exchange Online, calendar, sharepoint, Office clients, Skype for Business, Onedrive, etc. You're right that there are no ways to order VMs in there. You can purchase 10 licenses decently cheaply though, count on ~20$ a head for E3 licenses probably. These include basically everything in O365 (...maybe.... not Skype f Business and Project/Visio(??)), giant mailbox space, 1 TB per user on OneDrive + windows CALs for those users on your local AD if I'm not mistaken. It's a bit of work but to make it all work perfectly, but it's peanuts compared to hosting your own comparable mail servers etc If you're really lucky you can make them all work smoothly using only the tools Office365 provides and a few local machines (this won't happen). You could use OneDrive as a file server for each user. This used to be 'fraught with peril' but it's gotten a little better. Usually it's now seen as 'probably not the best idea' (but we've got the licenses anyway so it's free!!). But look into the services on offer, there isn't really a good hosted Excel for power users but most people can deal with the online clients. There's pretty good collaboration stuff in there, hosted Exchange works fine out of the box for smaller companies. Skype for Business is decent to good (depending on your current PBX) and Teams is pretty neat, especially by Microsoft standards. And you get sharepoint, which isn't cool at all but can be useful+Onedrive, which is basically also sharepoint but reskinned and slimmed down. You'll need to supplement this with local machines to use it all and a little local or cloud hosted back end at least. Then: the azure pricing calculator. https://azure.microsoft.com/en-us/pricing/calculator/ This is what you need for estimates of what VMs and database hosts and other infrastructure cost. You can create basically anything you can make on iron. However! You can't plug in your little 5V USB LED into one of those servers and sit in the room quietly reading by the white noise of the fans. This is overlooked too easily and imo a major disadvantage of full-cloud environments. Anyway in Azure you can make a lot of ready made pre-fab stuff as well, frankly if you need something that isn't in there in some way that would surprise me. But you're right. You're looking at a little more work now because 1) you have to figure out Azure itself 2) you need to get those users authenticated or in another directory, 3) you have to recreate what you had after. It isn't super hard to make a little VPN tunneling device, a virtual network, stick some file/script servers in there and go. But I don't think it's the best way of going about it. If you're going to replicate your old stuff exactly in a cloud environment, the only thing you move is hardware responsibilities. You're still going to patch it, fix whatever bugs, all the rest is still your problem. Also, probably going to be noticably more expensive (though the option to turn off non-critical hardware outside of working hours is really easy to use now, the portal will even ask/suggest you turn off your stuff to reduce bills). I'm not too up to date on the whole 'so macro-heavy they need a server' data processing thing, but there might be something that works perfectly for that. I know Microsoft like to mention PowerBI a lot and it's something to do with data but that's as far as my expertise goes.
|
|
#
¿
Jul 6, 2017 17:06
|
|
|
Tapedump posted:You’re right about their PST sizes, mostly. I see how that would be a big fat problem given how you describe profiles roam... e. woops missed a page there. point stands but I'll try to add some actual advice later IMAP is a genuine concern yeah but PSTs are arguably worse. At least, in my experience, might be overreacting but is a measurable percentage of your time related to rebuilding corrupt mail stores? It will once you start roaming profiles on PSTs. That PST and PTSD are nearly anagrams is one of those cosmic coincidences that make me believe the Intelligent Design people have a point Old Binsby fucked around with this message at 21:02 on Nov 24, 2017 |
|
#
¿
Nov 24, 2017 20:06
|
|
|
Matt Zerella posted:At the very least Exchange will be gone. Which would own. This is a reasonable position if you feel email sucks and I agree that it should get its descent into obsolescence over with already. Until then I'd much rather deal with Exchange 2019 (which the SaaS platform will also be based on) than the exim/postfix/dovecot/domino(?)/GoogApps alternatives around. though, while you're waiting for email to die keep in mind people (such as myself) are still out there herding farms of FoIP (yes, Fax-over-IP) gateways. I'm under 30 and I've spent time implementing literally Telex (over SIP)
|
|
#
¿
Nov 25, 2017 01:08
|
|
|
If Open Source is meant to imply free (of charge) that’s tough because for zero dollars you will always have to make compromises. I might be wrong about that but if free is what you need I didn’t hate the (closed source) free version of Veaam for what it was. Not perfect, but fine for smaller companies as long as you’re virtualizating using hyper-v/vmware. I have to say, a cryptolocker/virus hitting is to my knowledge the easiest shortcut to a blank cheque for fixing backup/monitoring properly for once, or at least relaxing cost and supplier restraints. Seems like that didn’t work in this case, I wonder what will
|
|
#
¿
Dec 20, 2017 19:15
|
|
|
NevergirlsOFFICIAL posted:
I like that one. Going to da cloud  is usually more expensive than small clients expect it to be and invariably more expensive than rolling your own Exchange/Sharepoint/SBS/postfix/dovecot/exim/DIY-setup. I don't know where the optimism about cutting costs even comes from but it's almost never justified. The same is true for larger companies (i.e >a couple thousand seats) who end up paying a lot due to sheer volume but they usually also have a dedicated licensing person around who shoujld see that coming (which they mostly don't). It's weird. I don't hear Microsoft sales people say "hosted services === cheaper" ever these days but the idea is quite persistent. Other topics I wrote words about on an internal blag in the past -How to set up a phishing campaign to train your employees at recognizing really obvious scams ... in under 30 minutes (subtitle: 'are you ready to face that people will click anything') -How to get every admin to embrace automation (Tough, I have had limited success by asking each to find something simple they do periodically and then teaching them all the same basics and adding specifics individually on how to automate their thing away after. Once they know the sweet taste of victory over boredom, repetition, their professional demise a lot are motivated enough to keep going with self study/ripping off stack overflow -Thought experiment: you're run over by a bus. How do you make sure your company IT survives this inconvenience? -Finding out who is uploading a Sponge Bob bit torrent through the company WLAN - a quick start in network monitoring -Locking down company laptops so Henry's nephew doesn't install a bunch of bit torrent stuff again - a primer in device hardening and compliance -
|
|
#
¿
Feb 14, 2018 07:22
|
|
|
Thanks Ants posted:I think the Cloud = Cheaper argument gets overdone and is too simple. But Exchange Online is cheaper (at least at the numbers that would qualify as a small shop) than doing it yourself properly - e.g. distributed data centres, some form of message filtering, a smarthost to ensure your sender reputation is always good, staff time to patch the servers, storing the mail databases on something durable etc. You're right I'm not arguing agains the economics of scale or in favor of running your own stuff on ancient hardware in a corner of the office. I just thought it funny that for some reason a lot of middle managers seem to have the idea ingrained that moving to cloud stuff means instant cheaper everything, even or especially in the short term. Thinking about it, the problem is probably that while you don't have to buy servers up front, of course you do still need to pay for licenses. Everything else is gravy: don't have to spend hours configuring a network, databases, proxies, load balancers, backup, database copies, virus scanning, spam filtering and so on. That's all very nice but the only cost that gets cut when you move is at the very front and it's a once-a-decade expense anyway. Whatever all of the rest costs is baked into the second item, the licenses, which is suddenly payable per month or year instead of sporadically at the end of a long hardware lifecycle process. I guess I'm answering my own question but no matter the gains in other areas, the total on the first monthly or yearly bill is usually higher than was expected.
|
|
#
¿
Feb 14, 2018 15:58
|
|
|
Rick posted:I'm incompetent and do not want to run SBS anymore. Until I started reading this thread I wasn't even aware SBS was still a thing. Goondolences.
|
|
#
¿
Feb 14, 2018 17:12
|
|
|
Terminal posted:Take all the roles & products that Microsoft normally advises segmenting between VM's / servers. Now mash them all together on whatever low-end Dell/Lenovo/HP server the MSP sold them running everything on a single RAID5 array. It's a wonderful product and easy to implement too. Take any AD/Windows Server/Exchange/Sharepoint performance or security guideline you know of. Do the exact opposite. Eventually, you'll have an SBS up and running. Collocate domain controllers with Exchange, stick a sharepoint farm in there, make sure it will run on any hardware post- 80386 and last but not least do everything via wizards. No need to let those pesky administrators have shell access (disclaimer: it has been a while since I last saw an SBS server so disregard this if any of the above has been fixed) (I'm pretty sure none of the above has been improved on though )
|
|
#
¿
Feb 14, 2018 20:07
|
|
|
GreenNight posted:PDQ Inventory is pretty great. This is the way to go, much easier learning the vagaries of system center once it’s setup correctly. Probably you’re going to break it at some point, I know I did, but knowing where you need to end up makes doing the second installation yourself much more doable
|
|
#
¿
Feb 25, 2018 09:37
|
|
|
MF_James posted:oh man I never saw this thread before, so here's an abridged version of a conversation I had with my favorite client. I love the ‘we’re losing a massive amount of $ per second fix it!!!!!’ thing. It’s always inflated bullshit and there’s this weird relation where the dollar amount seems to be proportional with the amount of penny pinching over your measly hourly rate after you fix it (which should be peanuts compared to the massive amount of money they’re supposedly making every single hour during normal operation)
|
|
#
¿
Mar 17, 2018 00:25
|
|
While our systems are down we are losing 100k AN HOUR!!!! FIX IT
Well, your drobo storage took a poo poo because, well I dunno, it's a black box and you refused to buy a support contract, luckily we get a free 30-day trial so I am working with them.
Thanks for all your help!
|
GreenNight posted:We're too loving cheap. I'll keep sniffing and using rubbing alcohol. i googled it and still have no idea what NRA sticks are, would you enlighten me? stickers? do people put stickers on company laptops??
|
|
#
¿
Mar 20, 2018 21:24
|
|
|
thebigcow posted:I haven't seen anything other than a mess of scripts run by task scheduler. I’m not a dba by any means but isn’t this what the sql agent is anyway? Maybe it shouldn’t be but any place where they let me touch the databases (poorly run small shops) this was basically what I thought I saw
|
|
#
¿
Apr 1, 2018 19:45
|
|
|
Digital_Jesus posted:I mean let them make it an IT problem. rip your home dir
|
|
#
¿
Apr 19, 2018 15:09
|
|
|
or just say gently caress 'em and don't allow those commie bastards to buy your poo poo anymore (except for the actual former commies who don't care either)https://forums.warpportal.com/index.php?/topic/235548-important-notice-regarding-european-region-access/ posted:Due to the changes of our company's service policy for the European regions, we are saddened to bring you news that, all games and WarpPortal services to the European regions listed below will be terminated on May 25th, 2018.
|
|
#
¿
Apr 26, 2018 23:13
|
|
|
Potato Salad posted:I mean, if you want took at a market 5% larger than the US and think "nah they're commies" because you can't be arsed to pony up some dev cash, I hope you're comfortable about your niche right up to the point someone who did pushes you out domestically. I agree 100% but quoted that because the shortsightedness and lack of business awareness made me laugh. Plus I never considered geofencing an international customer base, which is kind of original I think 'Make a reasonable eula and fix our business and software so we can be transparent to customers about their data and be GDPR compliant? Nah, let's just ignore 550 million potential users and boot the ones we already had' 
|
|
#
¿
Apr 27, 2018 01:14
|
|
|
Potato Salad posted:Ah okay gotcha we might, especially in the real businesses group you mention. They’ll be forced to do GDPR properly due to their size and they have the capacity for it. I think it’s more likely that the second category you mention is still going to be there and equally big, flying under the radar. There’s so many of those that they’re only going to get in trouble after some security incident arises because the chance of running into proactive auditing is so slim. pretty much the way things are currently, except the consequences of not doing things Right are worse
|
|
#
¿
Apr 27, 2018 10:27
|
|
|
I never had much trouble using WMI for that purpose so neither should a scanner tool. In a rush looking for a single kb lazily using this also works if you’re forgetful about WMI like me systeminfo | select-string ‘KBxxxx’ Don’t know how good/bad all that is in terms of the supercedence issues you’re running into because I don’t know what those are. Do you mean the scanner nagging because it doesn’t know Hotfix A, which you didn’t install, was rolled into Monthly Update B, which you did? Nessus is quite good in that regard in my experience
|
|
#
¿
May 3, 2018 08:14
|
|
|
Wrath of the Bitch King posted:Scanners are typically looking for compliance with CVEs. A fix provided can manifest in potentially hundreds of KBs, but in my experience scanners will choke if the KB produced in November of 2017 for a CVE in November of 2017 isn't present even if a January 2018 KB that encompasses the fix is there. I'm not sure why these scanners (at least the ones I have experience with) aren't looking at dll file revisions; most of them are looking for strings in the registry for this stuff, which is a special kind of dumb. I think we're talking about the same thing in that case. It can definitely be a hassle to make sense of a mile-long list of irrelevant 'critical' vulnerabilities. Which tool does this best, I don't know but I've only ever used nessus seriously enough with results that I could work with to recommend it.
|
|
#
¿
May 3, 2018 21:13
|
|
|
there’s probably a shorter, better version of that guide on serverfault.com though
|
|
#
¿
May 14, 2018 15:44
|
|
|
Be sure to check your printer supplier too, I've spent days on the phone with HP when Canon was to blame and the reverse. Edge isn't great but it's not even close to hardware suppliers in terms of shittiness
|
|
#
¿
May 27, 2018 15:08
|
|
|
wolrah posted:How is it that printer drivers can still manage to suck so hard? I don’t know but in general any hardware interfacing with users and outputting digital data is hard. I’ve had the most insane issues with drivers, i/o errors the past 5 years professionally even while actively avoiding hardware. all sorts of kit from bar code scanner drivers to analog or sip phones, conference a/v, fax machines that send/receive copies digitally, smart card readers, scanners, copiers, isdn-sip session managers, fingerprint readers, touch screens, pen input devices, midi controllers, audio mixers, buttons on an atmel iot dev kit, (...) keyboard and mouse are the only more or less universally solved issues
|
|
#
¿
Jun 4, 2018 23:28
|
|
|
to be honest i reread that post i wrote and i think i’ve made one just like it not too long ago or i’m having unreal deja vu. it’s my one unresolved it job trauma dodge hardware, dodge proprietary drivers by small companies who appear to have no more than one dev and 0-1 test devices, double barrel roll for companies that are still willing to support winXP, dodge com ports, serial ports, analog ports, rj45 interfaces 5 pole DIN plugs, dodge all that is my advice i think
|
|
#
¿
Jun 5, 2018 00:46
|
|
|
Digital_Jesus posted:Deposit the printer in the nearest dumpster. Hand user a sketch pad and colored pencils. Tell them they are now the printer. this depends on tact and personal charm but if they don't bite engage the emergency option and tell them they'll be chief reproduction officer
|
|
#
¿
Jun 5, 2018 01:05
|
|
|
Digital_Jesus posted:Sir this is the HR department. hey the phone disconnected weird maybe call IT about this later hm no my phone broke i can’t call. ah well. nothing we can do! let’s hr something
|
|
#
¿
Jun 5, 2018 02:47
|
|
|
Defenestrategy posted:Just did and currently the printer works, but I'll know more in morning when I come in and magically the printer stops working again. humming the portal song irl
|
|
#
¿
Jun 6, 2018 22:45
|
|
|
also, kaspersky stumbled down the ‘good software’ lists your friendly neighborhood compliancy officer keeps fast. they’re now just a notch above ‘state sponsored terrorist hacker collective’ in parts of europe. They’re facing a similar situation in the US i guess but i’ve been replacing a few kasperskys lately
|
|
#
¿
Jun 21, 2018 23:47
|
|
|
The Fool posted:I would, giving out blatantly wrong information when you don’t know the right answer is one of the worst things we can do. to the defence of the IT guy, though I agree that answer was quite stupid, 'no can do' can mean a couple things. Maybe they don't know how to and they're being lazy asking a coworker. Maybe they're straight up lying so you go away - both of those pretty dumb and bad. But being busy or tired or whatever and not in a mood to argue turns 'won't (due to company policy)' into 'can't' without explanation sometimes. Dumb because if people find out you lied they won't believe you ever again but lovely senior backing/not knowing the magic words 'i can't decide to make an exception for you myself, you can take it up with my manager' lead to this.
|
|
#
¿
Jun 28, 2018 00:11
|
|
|
Sheep posted:Second RingCentral, never had any problems with them and we even did fax over SIP (yes I know) without problems for years. i love the abbreviation FoIP because it looks and sounds subtly wrong just like fax over IP is. I deployed Faxination Server hooked up to Exchange a few times when a job featured prominent fax usage. Faxination is a good name too because it's an awkward pun on the word fax and a positive emotion, which just doesn't work but the software mirrors this by being really awkward to install, use, maintain and it doesn't work out ever trying to provide 100% uptime in HA environments.
|
|
#
¿
Jun 28, 2018 23:48
|
|
|
NevergirlsOFFICIAL posted:I was someplace the other day that had physical fax modem on a physical server receiving actual faxes. When the time eventually comes to kill the fax servers usually only a handful users actually actually have a need for one. One Simple Trick Fax Server Vendors Don't Want You To Know: you can just stick a real fancy fax machine on their desk and be done with it. Get a nice one. They're more reliable, (old) people know how to operate them and you'll still save a bunch probably over licensing and servers and whatnot
|
|
#
¿
Jun 29, 2018 00:32
|
|
|
Internet Explorer posted:Or just use a SaaS faxing service like the rest of us who decided to stop dealing with that poo poo? sane ideas for sure. I tried but got tangled up in cloud regulation and compliancy red tape too complicated to deal with for just a couple users in dinosaur land (finance IT). You're right though.
|
|
#
¿
Jun 29, 2018 00:45
|
|
|
NevergirlsOFFICIAL posted:PRTG is good
|
|
#
¿
Jul 6, 2018 15:24
|
|
|
Jack the Lad posted:Update: After going back and forth with Office 365 support for a few weeks, this is mostly fixed (it's still happening for emails from our own servers but they say they'll fix those too in 2-3 days). par for the course on EOP related stuff. Answer: the spam algorithm machine learned something dumb. They gently prodded it so your domain (or all of G apps?) would not be seen as super spammy. Unless recipients on other services were having the same issues, in that case i'd suspect blacklists. Though because EOP uses a few commercial, varying, spam engines even that might not be a guarantee
|
|
#
¿
Jul 27, 2018 05:38
|
|
|
people talking to each other....... hosed up if true
|
|
#
¿
Aug 1, 2018 13:59
|
|
|
carlcarlson posted:We couldn't implement IM at my old place because we couldn't get people to stop saying dumb poo poo in emails, that would then get turned over in a lawsuit. God only knows how e-discovery would have worked with chat logs. I imagine we'd have to turn over the entire chat log between employees, which would inevitably include them talking about wanting to see the new girls tits or whatever other terrible poo poo that was worse than what they already said in email. tell them very clearly everything that you are able to(/required to if that's the case) log any form of communication using their company device. Anyone unable to fathom the stupidity of things like discussing the new girls tits deserves the eventual shenanigans that will ensue if they send it to the wrong person or accidentally post something along those lines to a team group chat, the last of which I've seen happen with a phone jockey once. That guy had to go but no incidents since, it's been a couple years now. Skype f Business can do this at the start of every IM session for instance.
|
|
#
¿
Aug 2, 2018 13:47
|
|
|
NevergirlsOFFICIAL posted:subject line: ask your tam it'll be a thread full of tams
|
|
#
¿
Sep 5, 2018 23:27
|
|
|
|
| # ¿ Apr 23, 2024 17:09 |
|
|
pixaal posted:You can set the OU to block inheritance if you want to block all other GPOs. This is good for testing, but someone made a mistake in designing the OU structure if you need to use it heavily in production. Sometimes you just inherit a giant mess from the person before you and need to make it work while you figure out how to do it properly with as little downtime as possible. (This can be a nightmare and not worth doing if there are a bunch of other things reading AD and using the OU to determine what permissions users should have, but this is unlikely in a small shop probably 1-2 applications that'll need to be updated) ‘an insane man that gave me PTSD’ posted:‘an object can only be in one OU. so we can’t have a machine be both in a laptops and Finance and w. europe OUs at once. We needed to drill down - lowest common denominator - user based policy’ the above quote was in ‘documentation’ — top level AD schema objects filled w/text — someplace I worked a few months in the trenches doing damage control So the newbs wouldn’t find it and abuse its power (he actually told me to look there on the way out, cursed man)
|
|
#
¿
Nov 8, 2018 18:26
|
|