|
Wrath of the Bitch King posted:I really want to get into it too. My problem with it, from a high level, is adoption; it's all well and good to get things running using the technology, but once we get there and start growing we'll need to hire people that are familiar with it. I can only imagine that that's a nightmare as it stands. I've never seen a resume with DSC on it.
|
#
?
Sep 29, 2017 20:07
|
|
|
|
| # ? May 26, 2024 07:16 |
|
|
Tab8715 posted:Has anyone gone through their company's Blind? It's mostly bitching about indian poop habits, bitching about ludicrous pay not being ludicrous enough, perpetual layoff rumors, and peace out posts.
|
|
#
?
Sep 30, 2017 02:12
|
|
|
Sickening posted:The employees are going to be notified of the suspicious behavior today. We are going to let them know that if they would like their profiles exported and systems re-imaged that its available and ready. The mdt deployment with profile extraction is tested and working and only takes laptops about 25 minutes on average. How did this go? I can see this starting a loving IT riot
|
|
#
?
Sep 30, 2017 02:16
|
|
|
Paladine_PSoT posted:It's mostly bitching about indian poop habits, bitching about ludicrous pay not being ludicrous enough, perpetual layoff rumors, and peace out posts. You work for MSFT?
|
|
#
?
Sep 30, 2017 02:22
|
|
|
Can I ask for some advice and your experiences about leaving a full time role after a short period of time? I'm currently in a management role in a small consultancy and have been in the role for only 10 months. Due to the low pay and long unpaid hours it hasn't worked out so I'm looking for other options but I've never left a company after a short period like this with my average stint generally being at least 3 years. Any words of advice on how I can leave without burning all my bridges?
|
|
#
?
Sep 30, 2017 03:35
|
|
|
BaconBucket posted:Can I ask for some advice and your experiences about leaving a full time role after a short period of time? If you aren't getting paid then you aren't burning bridges saying you have taken another job. If it's not what was promised due to low hours then have a frank discussion with your manager when you turn in your notice. You can try asking ahead of time about ways to get more paid hours etc framed as career advancement. You aren't doing this out of the goodness of your heart. You are in it to get paid. H110Hawk fucked around with this message at 03:43 on Sep 30, 2017 |
|
#
?
Sep 30, 2017 03:41
|
|
|
As the poet Young Dolph says "Rule #1: always get the money first"
|
|
#
?
Sep 30, 2017 03:45
|
|
BaconBucket posted:Can I ask for some advice and your experiences about leaving a full time role after a short period of time? As a practice director at a mid sized consulting company who gets paid okay, I say get the gently caress out. I know my rear end is. Don't even care if my next stop is a management role at this point. 'This poo poo was a mega dumpster fire' in professional terms will suit my reason for leaving just fine.
|
|
|
#
?
Sep 30, 2017 04:04
|
|
|
BaconBucket posted:Can I ask for some advice and your experiences about leaving a full time role after a short period of time? You're leaving to take an opportunity for advancement that isn't available in your current role.
|
|
#
?
Sep 30, 2017 04:45
|
|
|
H110Hawk posted:If you aren't getting paid then you aren't burning bridges saying you have taken another job. If it's not what was promised due to low hours then have a frank discussion with your manager when you turn in your notice. You can try asking ahead of time about ways to get more paid hours etc framed as career advancement. My reluctance about leaving comes from the fact that they did give me a pay raise about 2 months ago due to my previous manager leaving but it's still peanuts for the long hours and level of stress. I understand you guys are right and I can foresee a really tense notice period. Has anyone got any tips on dealing with petulant/egotistical management during this period?
|
|
#
?
Sep 30, 2017 05:38
|
|
|
BaconBucket posted:My reluctance about leaving comes from the fact that they did give me a pay raise about 2 months ago due to my previous manager leaving but it's still peanuts for the long hours and level of stress. I understand you guys are right and I can foresee a really tense notice period. Has anyone got any tips on dealing with petulant/egotistical management during this period? Take the high road, kill them with kindness, stop working unpaid hours, drink to excess.
|
|
#
?
Sep 30, 2017 05:48
|
|
|
I found my favorite job posting. Qualifications / Certifications - REQUIRED Bachelor’s Degree in Computer Science or related field (4-year Degree) Minimum 10 years Windows Server Operating System Experience Microsoft MCP & MCSA Certifications VMware Certifications HP Certifications (Server and/or Storage) Cisco CCNA or equivalent Job Type: Full-time, Contract Salary: $20.00 to $25.00 /hour Required education:Bachelor's Required experience: Computer Server & Desktop IT Support: 6 years Required language: Excellent English Required licenses or certifications: WA Drivers License Microsoft Certified Professional (MCP) Cisco Certified Network Associate (CCNA) VMware Certifications HPE Certifications This is in an expensive suburb of Seattle, BTW. And no benefits. Edit: Oh, missed this - Ability to perform 24X7 Maintenance/Support ON-Call & ON-Site services for customer contracts (monthly and weekly schedule onsite visits). Peachfart fucked around with this message at 05:51 on Sep 30, 2017 |
|
#
?
Sep 30, 2017 05:48
|
|
|
"but you'll work 60 hours a week and overtime makes that $25 super valuable!! You'll be making more than your manager!"
|
|
#
?
Sep 30, 2017 06:03
|
|
|
Nitramster posted:Hello Thread, let me introduce myself: My name is Nitramster and I work at Geek Squad (get your punches in now, please) You're roughly where I was - I started in Geek Squad in my late 20s, was there for almost 5 years - moved up to 'DCI' for about 6 months before I got a real IT job. First thing: You do not want to be a Geek Squad supervisor, trust me. It's way too much work for one person, highly stressful, and grossly underpaid for what is expected out of you. There's also not much upwards mobility within Best Buy from there unless you are good at and want to learn more on the sales side. The most valuable thing you get out of Geek Squad is learning how to deal with users, particularly the stupid/stubborn/difficult ones. It sounds like you've already got a handle on this, so you aren't going to learn much more there. There's not enough time or resources available in the store to get actually useful tech skills beyond the basics of how to switch out hardware or repair very basic Windows problems - everything else gets shipped out, connected to AJU, or flattened/reimaged. My store's Mobile manager (who was also in GS for a while before that) got a job at the MSP in the office next to our store, and convinced me to apply. Within a month, I had FINALLY gotten out of retail, and was making about $10k more a year as a tier 1 helpdesk drone. Since then, I moved up to Tier 2, and then over to the NOC; I really hate myself for waiting so long to find something outside of retail. There's no reason to feel like you need to get certs before leaving - Helpdesk is an entry-level job almost everywhere, if you interview well and show you know how to find answers to stuff, you should be good to go. I still don't have any certs (note: YMMV here, but there's no reason to wait before papering the town with resumes). Having some time in Geek Squad can be helpful with entry level positions - it doesn't mean much on the technical side, but a lot of people have trouble handling users, you can use a few examples of that to show your interpersonal and de-escalation skills. I do agree with you that as far as retailers go, Best Buy isn't that bad - the health insurance and other benefits were decent, and they don't gently caress around trying to get free work out of their hourly people. But nearly anywhere else will be less stressful, provide more opportunities to learn, and pay more. The sooner you start looking, the sooner you can really launch your career.
|
|
#
?
Sep 30, 2017 07:49
|
|
|
Tab8715 posted:You work for MSFT? Err, yes. You're in my skype friends list
|
|
#
?
Sep 30, 2017 10:55
|
|
|
Ok so I'm starting to build a devops lab with an aws node, Jenkins, ansible, bash for Windows, GitHub, all that good stuff. Is completing a few lab projects good enough to get through the devops interview door? What should I be mindful of keeping or documenting for resume fodder? Do I just write a bunch of scripts with documentation and link my GitHub to prove experience? Basically if you were conducting an interview for a devops role and a systems admin walks in, what can they show you to lock in the job?
|
|
#
?
Sep 30, 2017 15:17
|
|
|
MANime in the sheets posted:
Amen. I'll be updating my resume today. If anyone in the L.A. area wants to network, send me an email: Nitramster at gmail. I'd be happy to buy any of you a beer.
|
|
#
?
Sep 30, 2017 15:58
|
|
|
Judge Schnoopy posted:Basically if you were conducting an interview for a devops role and a systems admin walks in, what can they show you to lock in the job? This is a question I would also love to know the answer to.
|
|
#
?
Sep 30, 2017 17:04
|
|
|
Walk to the whiteboard, and in one motion of the arm, draw a perfect circle.
|
|
#
?
Sep 30, 2017 17:22
|
|
|
Judge Schnoopy posted:Ok so I'm starting to build a devops lab with an aws node, Jenkins, ansible, bash for Windows, GitHub, all that good stuff. It all depends on on interviewer view of devops as it an mean a whole range of things to different people. I think generally you'll want to demonstrate your ability to automate the poo poo out of your environment. So learn how to spin up a vm, configure it, release code to it and test the deployment. Be able to explain the tooling you used to do the above, and have some basic knowledge of the other tools you could have used. That said the tech test of my current jobs consists of being able to ssh into a red hat box fix the listen address in httpd and enable the service.
|
|
#
?
Sep 30, 2017 19:26
|
|
|
fluppet posted:That said the tech test of my current jobs consists of being able to ssh into a red hat box fix the listen address in httpd and enable the service. I can do this. Hire me.
|
|
#
?
Sep 30, 2017 21:17
|
|
|
I'm sure Vulture Culture can follow up with a much better answer, but I think one thing people want to see to distinguish a "DevOps" person from a run of the mill sysadmin is the mentality. Are you comfortable with the idea of automation and disposable instances? Or are you stuck in the mentality of individually named servers that require special care and feeding and SSH-ing into them and making one-off configuration changes? That being said, there are plenty of little no-name companies desperate to hire "DevOps" people and would be happy with someone that has a SysAdmin background that is familiar with the basics of scripting, configuration management, and AWS/Azure. Make sure you also are familiar with Git. A familiarity with Jenkins or Teamcity and some CI concepts is going to make you even more compelling.
|
|
#
?
Sep 30, 2017 23:24
|
|
|
Paladine_PSoT posted:Err, yes. You're in my skype friends list  yes I remember now
|
|
#
?
Sep 30, 2017 23:48
|
|
|
Judge Schnoopy posted:Ok so I'm starting to build a devops lab with an aws node, Jenkins, ansible, bash for Windows, GitHub, all that good stuff. From a big SaaS company that's getting bigger, I would want experience with change control, config management, monitoring, and application security. You've got the first two, but adding centralized logging and managing iptables/selinux with those tools would be a huge plus. You can also roll your own gitlab instance and do encrypted S3 backups out of the box now Roargasm fucked around with this message at 00:37 on Oct 1, 2017 |
|
#
?
Oct 1, 2017 00:35
|
|
|
Wrath of the Bitch King posted:I really want to get into it too. My problem with it, from a high level, is adoption; it's all well and good to get things running using the technology, but once we get there and start growing we'll need to hire people that are familiar with it. I can only imagine that that's a nightmare as it stands. I've never seen a resume with DSC on it. I've been looking into DSC this week a bit and my biggest problem is finding resources about it. All the stuff I've seen is "check this out, you can ensure a file exists. This is how you set up a pull server." and that's about it. I guess this is my way of asking if anyone has good books or sites on DSC
|
|
#
?
Oct 1, 2017 02:24
|
|
|
myron cope posted:I've been looking into DSC this week a bit and my biggest problem is finding resources about it. All the stuff I've seen is "check this out, you can ensure a file exists. This is how you set up a pull server." and that's about it. microsoftdocumentation.txt But yeah, I don't really hear much about anyone doing anything with it, and most larger shops are already using a well established platform like chef or ansible.
|
|
#
?
Oct 1, 2017 02:46
|
|
|
Puppet has a module for using it as a framework to manage DSC settings, which seems... cute? I like the idea of leveraging DSC since it's there anyway, but I haven't wrapped my head around it.
|
|
#
?
Oct 1, 2017 04:20
|
|
|
Remember all that poo poo that I brought up that some people reacted negatively to? That's all Devops-y stuff, and it was the sort of stuff I was looking for when I was hiring people at my last job. Big things I'd look for: Pets VS cattle, thinking about how to design deployments to be redundant and individual instances to be not a single point of failure. Infrastructure as code(Terraform/Ansible/Salt/Chef/Puppet in that order, I don't think chef and puppet are super compelling anymore). This also bleeds into permission models in AWS with things like IAM roles. Familiarity with docker and some awareness of schedulers, big boy in this room is kubernetes. AWS, it's what most people use, and it's a common point of reference for comparing to Azure/GCP Metrics, and monitoring, nagios and some of the older monitoring tools are black marks to me here. Solar winds is trash. The new hotness is time series databases and not caring about individual nodes. If something hangs a health check will pick up the instance failing and something will re-provision it. Centralized logging, ELK, splunk. (I prefer stackdriver but it's rarer then the other two) Linux mainly, there are some windows Devops jobs but rarer, and mainly focused on building. This is more Ops focused then Devops focused but most places are not looking for you to write tools, you are more of a cloud engineer then a developer.
|
|
#
?
Oct 1, 2017 04:42
|
|
|
Is pets vs. cattle merely a physical/virtual host or VM bring a singe point of failure? That part of the equation but I've got a gut feeling there's more to it than that.
|
|
#
?
Oct 1, 2017 04:52
|
|
|
Tab8715 posted:Is pets vs. cattle merely a physical/virtual host or VM bring a singe point of failure? No. It's disposable poo poo. You spin up everything in a cluster. If something breaks gently caress it. Kill it and spin up a vm. Everything is automated so it does it without you logging in. Pets you gotta help them grow up.
|
|
#
?
Oct 1, 2017 05:25
|
|
|
Tab8715 posted:Is pets vs. cattle merely a physical/virtual host or VM bring a singe point of failure? jaegerx posted:No. It's disposable poo poo. You spin up everything in a cluster. If something breaks gently caress it. Kill it and spin up a vm. Everything is automated so it does it without you logging in. Pets you gotta help them grow up. I want to expand on this a little bit, because the very idea of cattle VMs is foreign if you're coming from an internal position at some not-technology company. If you've got stuff like maybe an on-prem installation of FileMaker and a handful of other little fish Windows things at some company with 100 salesmen/accountants/logistics people, it's basically useless to you and not relevant. Basically everything DevOps and what Punkbob just brought up is for when you're operating an online service of some sort. E-commerce, internet advertising, social media things, something like Datadog, whatever. This is where operating big online distributed systems that are actually the focus of your company. In this case absolutely everything should cattle because you'll have a bad time if your micro-services/compute nodes are pets and are not programmatically defined. These online services are not at all comparable to the monolithic Windows server applications might you throw in your server closest. Gross oversimplification: These micro-service applications are (should be) specifically written to be modular and stateless. You've got a message bus or publish/subscribe system of some sort and some backends that register themselves. These backends will take queries as they come, more backends means more requests can be handled. Since they're all stateless, mindless compute nodes its fine to raise and lower the number of backends according to real time use. Got a big spike of traffic every Monday morning? Sure, just make more backends to handle the load and then shoot them afterwards: only pay for what you use. I didn't fully appreciate or get this until I actually worked at a place where DevOps was applicable. My old small fry Windows shops just had no concept of it. I thought this was something important to mention because when DevOps is yelled at internal IT Windows admins, it just doesn't make any sense to the uninitiated. At least it didn't to me.
|
|
#
?
Oct 1, 2017 06:07
|
|
|
Yes, not using devops for my solidworks license server.
|
|
#
?
Oct 1, 2017 06:21
|
|
|
Methanar posted:I want to expand on this a little bit, because the very idea of cattle VMs is foreign if you're coming from an internal position at some not-technology company. If you've got stuff like maybe an on-prem installation of FileMaker and a handful of other little fish Windows things at some company with 100 salesmen/accountants/logistics people, it's basically useless to you and not relevant. This
|
|
#
?
Oct 1, 2017 07:39
|
|
|
Methanar posted:I want to expand on this a little bit, because the very idea of cattle VMs is foreign if you're coming from an internal position at some not-technology company. If you've got stuff like maybe an on-prem installation of FileMaker and a handful of other little fish Windows things at some company with 100 salesmen/accountants/logistics people, it's basically useless to you and not relevant. As a guy in a pretty small shop, this is so strange to me that I am grateful you expanded on it. I still don’t think I fully get it, but I understand it quite a bit more
|
|
#
?
Oct 1, 2017 14:14
|
|
|
Methanar posted:I want to expand on this a little bit, because the very idea of cattle VMs is foreign if you're coming from an internal position at some not-technology company. If you've got stuff like maybe an on-prem installation of FileMaker and a handful of other little fish Windows things at some company with 100 salesmen/accountants/logistics people, it's basically useless to you and not relevant. This should be added to the OP for sure. I'm going to use it to explain this to some of my friends in tech that think what we work on at my job is black magic.
|
|
#
?
Oct 1, 2017 15:04
|
|
|
Methanar posted:I want to expand on this a little bit, because the very idea of cattle VMs is foreign if you're coming from an internal position at some not-technology company. If you've got stuff like maybe an on-prem installation of FileMaker and a handful of other little fish Windows things at some company with 100 salesmen/accountants/logistics people, it's basically useless to you and not relevant. I'd also add that most windows admins already treat some compute resources as cattle, in particular client machines. If a client box gets a virus or a bad hard drive most places replace and reimage. It's a similar concept. In the cattle vs pet scenario with servers there is just a super set of events that will trigger something like a rebuild, or a scaling event. Maybe the nodes drive filled or the docker container locked up. I don't care because my orchestrator recovers from that. Now if there is a systemic issue I will look at it, and often I might debug node failures if I have the time so that I can see if it's a config issue, but I don't have to. My service is up and people are happy. Now yes if you are in a small windows shop pet vs cattle is somewhat useless to you, but infrastructure as code and config management aren't. At a high level it's the concepts introduced in AD GPOs but more explicit. I want all of my configs saved in git so that I can see a history of config changes, prevent config drift and have peer review on config changes. This is where I also drop in my note that I only brought any of this up because some asked what to study to get a "Devops" job because it pays $$$$. I don't want to be called a cloud shill again. And to ape Cato, this is where I add that if you think that SaaS isn't coming to eat your windows shop, and it's FileMaker installs then I think you're naive.
|
|
#
?
Oct 1, 2017 15:39
|
|
|
Punkbob posted:I'd also add that most windows admins already treat some compute resources as cattle, in particular client machines. If a client box gets a virus or a bad hard drive most places replace and reimage. It's a similar concept. you keep talking about config drift, which is something i have never seen anyone mention before, and no one else who does devops here talks about. Care to elaborate on that?
|
|
#
?
Oct 1, 2017 18:50
|
|
|
I guess what he means with config drift is people (more aptly named idiots) who start messing with config files and not just changing them in the central place (git) and then pushing it to all apropriate servers (if it's not already happening automatically). This means server 1 has a different config than server 2-20 making it a unique little snowflake and troubleshooting it a major bitch. In a cattle environment it's no biggy. Kill it and spin up a new one. In a pet environment you start womdering why something works on all but 1 server. Or even worse, why only one of your servers is working great and the rest is a pile of poo poo. No config management/enforcement is hell.
|
|
#
?
Oct 1, 2017 19:36
|
|
|
RFC2324 posted:you keep talking about config drift, which is something i have never seen anyone mention before, and no one else who does devops here talks about. Care to elaborate on that? Config drift is the idea that you can't keep people from making little changes and tweaks, and eventuality your puppet/chef/ansible/whatever files will be out of date and when you spin up a new vm nothing will work right.
|
|
#
?
Oct 1, 2017 19:37
|
|
|
|
| # ? May 26, 2024 07:16 |
|
|
Well, poo poo. I know in my puppet systems I have my systems refresh config from the central repo on a regular basis. You want to change a config, its gonna revert pretty quick, and if you need to tinker, you need to shut down the puppet agent, which will alert that something funny is going on.
|
|
#
?
Oct 1, 2017 19:40
|
|