|
Polio Vax Scene posted:Apparently a lightning strike hit Microsoft's San Antonio data center and has been loving things up for me all day. hosed a lot of stuff up today for us too. I sent an email out to the faculty in the morning once we figured it what was going on, so of course they just kept trying all day and sent loads of kids down to me when it didn’t work.
|
#
?
Sep 4, 2018 22:31
|
|
|
|
| # ? Jun 4, 2024 17:13 |
|
|
Someone is gonna make bank selling cloud business continuity planning to the myriad of poo poo companies who have only one plan: Panic
|
|
#
?
Sep 4, 2018 22:41
|
|
|
Crowley posted:We just started rolling out Cylance. Have to say I'm pretty impressed so far.
|
|
#
?
Sep 4, 2018 22:52
|
|
|
Partycat posted:Someone is gonna make bank selling cloud business continuity planning to the myriad of poo poo companies who have only one plan: It’s really strange how people got the impression that running in the cloud doesn’t mean that you still need to architect your stuff in a way that can cope with data centre outages - none of the platforms hide the fact that there are multiple regions that you need to choose from, and all the presentations I’ve seen delivered have put the onus on the customer to build the resilience and security. Is there a bunch of marketing being delivered to CEOs that says the cloud solves all the problems, and they veto any plans by ops teams to deploy in multiple regions?
|
|
#
?
Sep 4, 2018 23:11
|
|
|
Thanks Ants posted:It’s really strange how people got the impression that running in the cloud doesn’t mean that you still need to architect your stuff in a way that can cope with data centre outages - none of the platforms hide the fact that there are multiple regions that you need to choose from, and all the presentations I’ve seen delivered have put the onus on the customer to build the resilience and security. We had a decent outage on our main website due to a us-east-1 outage a while back. Turns out our site that's big enough to pay for a superbowl ad for doesn't really need cross-az or cross-region availability!
|
|
#
?
Sep 4, 2018 23:21
|
|
|
VSTS was pretty badly affected, even for European customers, so maybe MS should take a page from their own book and host global applications... globally?
|
|
#
?
Sep 4, 2018 23:35
|
|
|
Yeah if you’re consuming SaaS then there’s no excuse
|
|
#
?
Sep 4, 2018 23:43
|
|
|
You would think loving Microsoft would have some redundancy but it makes me lmao when a south central US datacenter outage is affecting services globally. We live in backwards land.
|
|
#
?
Sep 4, 2018 23:59
|
|
|
Crowley posted:We just started rolling out Cylance. Have to say I'm pretty impressed so far. Cylance is great if you host it on prem The cloud version is poo poo and you have next to no customization options for exclusions. It was so bad we switched vendors 8 months into our contract. It broke almost every single application we use in house, most notably Adobe on Mac... even though its functionality on Mac is limited as heck.
|
|
#
?
Sep 5, 2018 00:38
|
|
Cat Army 
|
Thanks Ants posted:It’s really strange how people got the impression that running in the cloud doesn’t mean that you still need to architect your stuff in a way that can cope with data centre outages - none of the platforms hide the fact that there are multiple regions that you need to choose from, and all the presentations I’ve seen delivered have put the onus on the customer to build the resilience and security. The cost benefit goes away when you build in resiliency. 99% of companies will just live with two or three outages a year since they are only a few hours long. Gotta buy more services and pay better monkeys to set it up “the right way”, and it’s just not worth hurting the bottom line
|
|
#
?
Sep 5, 2018 01:01
|
|
|
Bob Morales posted:The cost benefit goes away when you build in resiliency. 99% of companies will just live with two or three outages a year since they are only a few hours long. I mean its a matter of if downtime is acceptable for what you're building. E.g a car website can probably afford some downtime. The Google SRE book has a really good bit on what they call an 'outage budget', which basically boils down to comparing the cost of more reliability vs how crucial what you're doing is.
|
|
#
?
Sep 5, 2018 02:35
|
|
|
Polio Vax Scene posted:You would think loving Microsoft would have some redundancy but it makes me lmao when a south central US datacenter outage is affecting services globally. We live in backwards land. You do not want to be in the hot seat over service outages due to lightning strike 
|
|
#
?
Sep 5, 2018 02:37
|
|
|
Aunt Beth posted:Our IT security has FINALLY decided to get rid of McAfee and this is one of the potential replacement products. What have you been impressed by? We're currently in the "Discovery" phase. We've rolled it out to a number of servers and clients and are just monitoring and going through what to whitelist. So far I'm impressed by the things it detects, how it detect threats, and how little impact it has on a running system. Here's an example of why Cylance thinks a specific sys-file should be examined further.  The Iron Rose posted:Cylance is great if you host it on prem We're on cloud, and I don't quite get what you mean with customization options for exclusions. So far it's been pretty easy to whitelist a file or a group of suspicious files for a single device, one or more zones (groups), or everyone. Also: We only support PCs, which is nice.
|
|
#
?
Sep 5, 2018 08:07
|
|
|
Cylance has been pretty good, but it took a while to get there. Only recently did they add more reasonable hash-based file exceptions, and its still not global which is dumb af. I have had no problems with its functionality on mac at all. What issues did you have with it Iron Rose?
|
|
#
?
Sep 5, 2018 20:09
|
|
|
We had persistent but inconsistent issues with exclusions applying properly across device zones. Hash-based whitelisting tended to work fine, but every other week or two it'd flag, for example, the adobe crash reporter as suspicious, quarantine it, and break the application until it could be redownloaded. Their reporting options were nonexistant - I had to manually queue a giant CSV of all encountered threats every week and do some powershell table wizardry to output something reasonable for our helpdesk and security people to respond to incidents. Usually it'd be a legitimate application which we'd flag as not a threat until the next time Cylance flagged and quarantined it anyways. Support was slow and not terribly responsive. It was, however, extraordinarily effective at catching legitimate threats, I can't think of more than one or two cases where something slipped by it - our issues at my previous company were mostly with it disliking our internal applications. The script control function in particular was very aggressive, and did not always respect whitelisting. Most notably, despite the fact that script control wasn't meant to function at all on OSX in the first place, having it enabled for a device zone with OSX machines in it would interfere with JAMF deployments, etc. We were using threat defense though, so our administrative options were incredibly limited to begin with. If your place springs for the whole package, it's not a bad option. This was about a year and change ago too, so maybe they've made it better.
|
|
#
?
Sep 5, 2018 21:31
|
|
|
According to Comcast our newly acquired location needed a new cable pulled from the pole to the building. So the tech that was installing the service put in a work order to have it done at a later date, and put in a booster to bring the signal level up a bit until they replaced the cable. Well another tech shows up today to get this done. The tech calls ME and asks if I know which street pole their equipment is on. I actually thought he was joking at first..... he wasn't.
|
|
#
?
Sep 6, 2018 19:49
|
|
|
I'd be all "the one out front."
|
|
#
?
Sep 6, 2018 19:52
|
|
|
Oh god... Got another call from him.. "I found one cable out there and I unhooked it, but your service is still working, so that wasn't it... I am going to call my boss" Good thing we have a backup connection, because it looks like we are going to be using it.
|
|
#
?
Sep 6, 2018 19:59
|
|
|
stevewm posted:Oh god... Got another call from him.. "I found one cable out there and I unhooked it, but your service is still working, so that wasn't it... I am going to call my boss" I mean, if they're pulling a new line in, that means they're unhooking the current one so, yeah, you're going to lose service for a bit.
|
|
#
?
Sep 6, 2018 20:01
|
|
|
MF_James posted:I mean, if they're pulling a new line in, that means they're unhooking the current one so, yeah, you're going to lose service for a bit. I should have elaborated a bit... It sounds like instead of the quick cutover they promised, its going to be a multi-day ordeal, likely with no service at all.
|
|
#
?
Sep 6, 2018 20:03
|
|
|
stevewm posted:I should have elaborated a bit... It sounds like instead of the quick cutover they promised, its going to be a multi-day ordeal, likely with no service at all. Sorry was just being an internet jerk about it, I figured you meant something like that.
|
|
#
?
Sep 6, 2018 20:04
|
|
|
They promised a quick cutover amounting to just minutes. Why do I still continue to believe telecom company promises?
|
|
#
?
Sep 6, 2018 20:10
|
|
|
stevewm posted:According to Comcast our newly acquired location needed a new cable pulled from the pole to the building. So the tech that was installing the service put in a work order to have it done at a later date, and put in a booster to bring the signal level up a bit until they replaced the cable. stevewm posted:Oh god... Got another call from him.. "I found one cable out there and I unhooked it, but your service is still working, so that wasn't it... I am going to call my boss"
|
|
#
?
Sep 6, 2018 20:15
|
|
|
stevewm posted:They promised a quick cutover amounting to just minutes. Why do I still continue to believe telecom company promises? Yep. I cancelled my personal Comcast account, pretty simple after I said I was getting gig speeds for $50/mo, and yes I had tested it. Then I get a letter with a bill for the month after I cancelled and it had auto paid. Thankfully they actually refunded it, I made sure I had a copy of the cancellation email in case they tried to deny it. Still going to keep an eye out for more charges.
|
|
#
?
Sep 6, 2018 20:19
|
|
|
I don't see why Comcast can't string a new cable from the pole, attach it to your building, run it into wherever your modem is going, get the ends terminated, do the same at the pole, and then swap it over. Once the new service is working then rip the old line out. Maybe that's too logical for a telco.
|
|
#
?
Sep 6, 2018 21:19
|
|
|
Thanks Ants posted:I don't see why Comcast can't string a new cable from the pole, attach it to your building, run it into wherever your modem is going, get the ends terminated, do the same at the pole, and then swap it over. Once the new service is working then rip the old line out. Maybe that's too logical for a telco. What interest whatsoever would Comcast have in providing uninterrupted service
|
|
#
?
Sep 6, 2018 21:22
|
|
|
Thanks Ants posted:I don't see why Comcast can't string a new cable from the pole, attach it to your building, run it into wherever your modem is going, get the ends terminated, do the same at the pole, and then swap it over. Once the new service is working then rip the old line out. Maybe that's too logical for a telco. That is what I was originally told they would do. But I get the impression the guy they sent to do it doesn't even know where their own terminations are on the poles.
|
|
#
?
Sep 6, 2018 21:33
|
|
|
Pissing me off today: Coworkers not reading PR descriptions or commit messages, then asking why I created a new template tag they think for some reason is identical to the old one we were using from a package. Reading comprehension issues aside, do you not see the reduction of the size of that template by like 50+ lines  It's really disappointing, because I thought my solution to the problem was rather clever and was hoping they'd notice 
|
|
#
?
Sep 6, 2018 23:43
|
|
|
BaronVonVaderham posted:Pissing me off today: Coworkers not reading PR descriptions or commit messages, then asking why I created a new template tag they think for some reason is identical to the old one we were using from a package. Maybe they never saw it as a problem. Therefore, you reducing something they never knew existed by 50 lines is unlikely to get noticed.
|
|
#
?
Sep 7, 2018 02:33
|
|
|
I’m in awe at the blunt honesty
|
|
#
?
Sep 7, 2018 02:46
|
|
|
Volguus posted:Maybe they never saw it as a problem. Therefore, you reducing something they never knew existed by 50 lines is unlikely to get noticed. Except it's highlighted in bright red in the diff on the PR page on github. It's pretty clear. It's also code that was in my previous PR that got the comment, "Why are we duplicating so much code, there has to be a better way but this is fine for now." So it's right in their face and is something that was directly requested a day ago.
|
|
#
?
Sep 7, 2018 03:41
|
|
|
If someone needs access to an entire department's email/calendars, there's something wrong, right? Right?!
|
|
#
?
Sep 7, 2018 15:16
|
|
|
Is free/busy broken?
|
|
#
?
Sep 7, 2018 15:31
|
|
|
Request came in: 'I need public internet traffic inbound to this VM' Me, a reasonable person: '  that VM is like 8 layers deep into our defense-in-depth posture, allowing unrestricted internet straight in from anywhere is like putting a screen window on a bank vault door'Requester: 'Well I need it.' OH OK
|
|
#
?
Sep 7, 2018 16:40
|
|
|
I'm aware this is really minor stuff, but it bugs me when people can't get the names of things correct. If you're talking about a product name, the name of your client, the name of a location - just get the capital letters, spaces, and whether it's plural or not correct. It doesn't take long and it looks quite bad if you've got the company name wrong of someone you want to work with.
|
|
#
?
Sep 7, 2018 20:36
|
|
|
Thanks Ants posted:I'm aware this is really minor stuff, but it bugs me when people can't get the names of things correct. If you're talking about a product name, the name of your client, the name of a location - just get the capital letters, spaces, and whether it's plural or not correct. It doesn't take long and it looks quite bad if you've got the company name wrong of someone you want to work with. Look buddy my Linskys AirPort is supposed to handle my spelling for me, ok?
|
|
#
?
Sep 7, 2018 21:20
|
|
|
I got a new computer. Clean windows install. Updated. SQL Server Management Studio won't install without some loving around, but ok, got it installed. Works fine. Then, later, I install Microsoft Office. Installing office completely breaks SSMS due to some sort of dependency. SSMS won't reinstall, won't uninstall. None of the tools that purport to deal with the situation work. So now I'm reinstalling windows, because Microsoft can't be bothered to make sure that their software won't break their other software. Creating the Windows install media on the usb key failed 3 times. The solution was to format the USB key as NTFS instead of FAT32, and to run the install media creator FROM the USB key rather than the computer.
Vim Fuego fucked around with this message at 21:43 on Sep 7, 2018 |
|
#
?
Sep 7, 2018 21:38
|
|
|
You guys ever experience that feeling the second that you decide that you're going to try and leave a job? Like you're doing all the same stuff but you look at it in a completely different way like you're putting in lip-service to not get fired until you find something else instead of actually trying to be succeed at your job. That's how I'm feeling today.
|
|
#
?
Sep 7, 2018 22:04
|
|
|
Thanks Ants posted:I'm aware this is really minor stuff, but it bugs me when people can't get the names of things correct. If you're talking about a product name, the name of your client, the name of a location - just get the capital letters, spaces, and whether it's plural or not correct. It doesn't take long and it looks quite bad if you've got the company name wrong of someone you want to work with. My lab top hard drive won't turn on
|
|
#
?
Sep 7, 2018 23:04
|
|
|
|
| # ? Jun 4, 2024 17:13 |
|
|
Vim Fuego posted:I got a new computer. Clean windows install. Updated. SQL Server Management Studio won't install without some loving around, but ok, got it installed. Works fine. Then, later, I install Microsoft Office. Installing office completely breaks SSMS due to some sort of dependency. SSMS won't reinstall, won't uninstall. None of the tools that purport to deal with the situation work. So now I'm reinstalling windows, because Microsoft can't be bothered to make sure that their software won't break their other software. Creating the Windows install media on the usb key failed 3 times. The solution was to format the USB key as NTFS instead of FAT32, and to run the install media creator FROM the USB key rather than the computer. Possibly not the issue and you may have already tried it, but the normal uninstaller for Office 2016 is hot garbage, you want to use the complete uninstaller: https://support.office.com/en-us/article/uninstall-office-from-a-pc-9dd49b83-264a-477a-8fcc-2fdf5dbf61d8 We ran into a bunch of issues with Office on some Surface Pros, regular uninstalls didn't do anything, this fixed it.
|
|
#
?
Sep 7, 2018 23:04
|
|
