|
The Iron Rose posted:I’m not even “open for work” on LinkedIn and I’m getting a bunch of recruiter spam. Chalked it up to being three months into my new SRE role. If you don't want to do them, that is one thing. If you don't feel like you should, don't limit yourself because you feel you should. The rest of the world isn't.
|
# ? Feb 13, 2021 19:53 |
|
|
# ? May 29, 2024 16:09 |
Management sucks anyways, I won’t try to move back into the track again unless I’m ready to acknowledge I don’t want to continue being a technical demigod or I got too dumb to keep getting better.
|
|
# ? Feb 13, 2021 20:46 |
|
The Iron Rose posted:I’m not even “open for work” on LinkedIn and I’m getting a bunch of recruiter spam. Chalked it up to being three months into my new SRE role. I was managing a department of 5 at age 20. Like Sickening said, if you don't want to do it, then don't. But if you think it's something you'd be good at and have a passion for, don't let lack of experience stop you.
|
# ? Feb 13, 2021 20:54 |
|
The Iron Rose posted:My favourite are the recruiters hitting me up for team lead, senior SRE, and management roles. I’m 25 my dudes, let me get a biiiit more emotional maturity please and thanks. Most of my managers had the emotional maturity of a toddler high on sweets so don’t let your age block you from trying new ventures
|
# ? Feb 13, 2021 20:59 |
|
Old job just reached out for a contracting opportunity to help migrate them to teams voice from a Cisco ucm. My be able to get $15-$20k out of it
|
# ? Feb 13, 2021 21:09 |
|
The Fool posted:Old job just reached out for a contracting opportunity to help migrate them to teams voice from a Cisco ucm. My be able to get $15-$20k out of it Can't even get a decent new boat for that money anymore! Congrats
|
# ? Feb 13, 2021 21:22 |
|
Internal recruiter chat seems to have went well for the SRE position, they want to schedule a talk with the hiring manager. Always excited to almost double my salary and not have to deal with poo poo tier health insurance for once in my life.
|
# ? Feb 13, 2021 22:41 |
|
Out of town, doing a migration. One user account is not mapping drives no matter what my guy does. I go start poking at it, I can't really understand why the GPO is not taking. Then I try to browse to the \\server\share and I get hit with a login prompt. Yup, she was not in the correct Security Group. Sigh.
|
# ? Feb 14, 2021 00:40 |
|
tortilla_chip posted:Take their money. We need to be able to sticky comments to the top of the forum.
|
# ? Feb 14, 2021 00:53 |
|
I may have landed an IT Director position (lol title inflation this is infrastructure manager at best) wearing nothing but semi wrinkled shirts and underwear. WFH covid interviewing owns so much.
|
# ? Feb 14, 2021 01:35 |
|
We might start a u2f pilot next year, I was eyeing yubi 5ci and 5c but Amazon reviews are kinda bad for the 5ci, lots of broken keys on the lightning side. Does anybody have a 5ci and if you do how sturdy it is? We do have an handful of ipads so lightning would help greatly.
|
# ? Feb 14, 2021 20:29 |
|
I've got a dumb question that I hope someone smarter than me can answer. I hope this is the right thread for it because I'm not sure. Previously, a guy managed the environment entirely. That person is no longer responsible for anything as they have been booted out. DNS records for the domain were all hosted by "his company." The registrar is Network Solutions, and the Name Servers are pointing to his Name Servers. I've determined after talking to several companies, that setting up DNS records on the registrar prior to switching to the registrar's nameservers to make a seamless transfer is nothing but a pipe dream. How much of an outage should I be estimating, if one at all realistically? In my mind, it's all unknowable. I change the Nameservers to the registrar, set up the DNS records as quickly as I can, and just wait for everything to populate. Network Solutions claims it takes 24-36 hours for the Nameservers to change, so theoretically it could be no downtime. Realistically that could be the normal hogwash of "it takes 24 hours" and it really takes 5 minutes and we are looking at a short downtime. I'm really out of my realm on this and I need an adult to tell me it's all going to be OK.
|
# ? Feb 15, 2021 16:19 |
|
Get a Cloudflare account, replicate your DNS zone in Cloudflare, change the name servers on the domain at the current registrar to Cloudflare. Then wait a few days and transfer the domain. I don't know any registrars that reset name servers on domain transfers. Registrar-supplied DNS is poo poo, unless you're using Route53 as your registrar.
|
# ? Feb 15, 2021 16:40 |
|
This is definitely possible to do with no downtime. I’m familiar with Google DNS and Route53, you should be able to set up records in your hosted zone and just wait till everything has propagated to point your nameservers at the new registrar. There is no restriction on having multiple nameservers from multiple companies. Registrar supplied DNS is indeed poo poo, use a real provider like cloudflare or one of the big three cloud companies. The Iron Rose fucked around with this message at 17:14 on Feb 15, 2021 |
# ? Feb 15, 2021 17:09 |
|
The key point here is that you can test this yourself once you have the records setup. Basically today you do: Host myawfulcompany.com Nslookup myawfulcompany.com This uses your local recursive resolvers and they lookup those nameservers from the global root servers, then ask them for the answer. However you can tell your lookup program to "start with this server" and in this case you would put in wherever you have setup your zone (myawfulcompany.com). Those are what you intend to update in the registrar. Host myawfulcompany.com ns01.myshittydnsprovider.com Now you can test it in advance and switch with confidence. Test every single record. Want to see a demonstration of how the global dns system works? Find a mac/Linux machine, or install dig for windows. 'dig +trace myawfulcompany.com' H110Hawk fucked around with this message at 17:55 on Feb 15, 2021 |
# ? Feb 15, 2021 17:52 |
siggy2021 posted:I've got a dumb question that I hope someone smarter than me can answer. I hope this is the right thread for it because I'm not sure. NS isn't the best but you can add all txt in one go, then hit MX and so on. If you have everything ready you are looking at it taking 15 mins to setup. Mostly giving the poo poo NS gui to update your changes. Setting a new nameserver prior is the best go, but a day to setup is bull and I would guess minimal downtime.
|
|
# ? Feb 15, 2021 18:12 |
|
ew, GUIs Set it all up in terraform and track changes in your VCS. You not only have a PR for every new record establishing who did it and why, but your changes will take less than 5min to apply via your provider’s API. Really though if you do this right the time it takes to create new records shouldn’t matter at all, whether it takes 5 minutes or 5 hours. Setup the records on the new provider first before updating your NS records and you’ll be fine. The Iron Rose fucked around with this message at 18:21 on Feb 15, 2021 |
# ? Feb 15, 2021 18:19 |
|
Thanks for all the replies, I think I have a better idea of what is going on now. There has been a lot of confusion around this because people told me things that weren't true, other people left out info until the last minute. I'm not looking to transfer domain registrar's, FYI. That might happen down the road, but the situation right now is that the end owner of this domain (it's a client, not mine) has no control over any of their DNS records. We need to add a TXT record for a verification for an important service, or important service might stop working. Person who actually controls the DNS records is an enormous douchenozzle and will not just add this record for us until we get poo poo figured out. I have no idea what Cloudflare or other DNS services cost for Nameservers. My entire experience with DNS is either with a registrar, or using Digital Ocean nameservers for side projects I'm working on. Submarine Sandpaper posted:How many subdomains are we talking? About six. Most of them are just MX records. The Iron Rose posted:ew, GUIs Terraform is way overkill for this (although I'm learning terraform recently, but for deploying Red Team infrastructure and it rules). If I could set up the records on the registrar before updating the NS records that would be great.
|
# ? Feb 15, 2021 18:45 |
|
that's a massively different problem set. you now need to get a lawyer involved because your previous employee is holding an existential part of your business hostage, which is illegal. Usually a threatening letter from counsel is enough here, if this guy is truly that stupid (and your business is small enough that there are only MX records), you're probably better off taking the downtime. Not being able to verify you own the domain or add DNS records is completely debilitating. The Iron Rose fucked around with this message at 18:57 on Feb 15, 2021 |
# ? Feb 15, 2021 18:54 |
I think on their GUI you cannot. Pointing it to another name server locks out advanced DNS settings. You should be able to get 6 subs and the MX records set in like 5 mins. Just hope your internet doesn't go out when clicking buttons. I've never been assed to see of I could manage my clients' network solutions stuff in any other manner so maybe it is possible via API.
|
|
# ? Feb 15, 2021 19:22 |
|
Cloudflare is free
|
# ? Feb 15, 2021 19:49 |
|
siggy2021 posted:Thanks for all the replies, I think I have a better idea of what is going on now. There has been a lot of confusion around this because people told me things that weren't true, other people left out info until the last minute. You can absolutely do everything you need before hand in Cloudflare or Route53. Just make sure it's in a new account properly owned, controlled, and payed for by company/person/entity who actually owns the domain / business. Cloudflare is nominally free for DNS (though you can spend plenty on them if you're big). Route53 is nominally free for any sort of small dns usage (don't get me started on AWS account best practices). Small note, technically the NS and SOA records mean nothing, they're not how DNS actually works when it comes to domains, once everything is setup you'll want the registrar to change the authoritative nameservers, this is not the same as setting dns records or transfering a domain registration. These are the things that a whois against the registrar returns.
|
# ? Feb 15, 2021 21:43 |
|
Has anyone actually seen a secondary site, failover, testing or disaster recovery in Azure or AWS with VMware or another hyper-converged provider like Nutanix? I've seen the marketing material but to hear from people on the ground if this actually works and delivers. My million dollar question, is could the replication be real-time or how close could I get? Gucci Loafers fucked around with this message at 00:21 on Feb 16, 2021 |
# ? Feb 16, 2021 00:06 |
|
Crosby B. Alfred posted:Has anyone actually seen a secondary site, failover, testing or disaster recovery in Azure or AWS with VMware or another hyper-converged provider like Nutanix? It can be just as replicated as you could get onsite. The issue is only how big the mpls circut and how much data you are averaging. With an express route we could replicate without issue but we also weren't replicating huge, changing workloads. Scope and scale is most important.
|
# ? Feb 16, 2021 00:36 |
Crosby B. Alfred posted:Has anyone actually seen a secondary site, failover, testing or disaster recovery in Azure or AWS with VMware or another hyper-converged provider like Nutanix? Yes. Azure Site Recovery works fine as long as you have the bandwidth to replicate and the correct number of process & config servers for the number of servers you’re protecting. Over the internet works just as well as an express route (private with private endpoints or over Microsoft peering) as long as replication is good you can have app consistency down to 15 minutes with multi VM groups. I think it’s 15 minutes and I’m posting from my phone and this loving MS documentation is rear end in a top hat. It might be shorter than 15 minutes.
|
|
# ? Feb 16, 2021 00:58 |
|
Crosby B. Alfred posted:Has anyone actually seen a secondary site, failover, testing or disaster recovery in Azure or AWS with VMware or another hyper-converged provider like Nutanix? Drawing that green line is what separates the "$LOL" engineering orgs from the $fine ticket jockeys. You can't beat the speed of light, if you can come talk to me privately because we're about to make all the dollars. From there - what's your goal? What's your application? If you want to pay VMWare to "make it happen" they will - the question is will you choose to afford the pricetag? If not, suddenly it doesn't work and everyone is mad at you. If you set it up but refuse to do real world tests it also won't work and everyone is mad at you. You MUST pull the plug and then start DR. If you start from "do_the_needful.sh" then of course it worked - that's controlled. Go have a datacenter tech pull enough power cords on your router to shut it off, THEN run your script. Have I seen it work? Yes, absolutely, but we spent exactly $0 on any enterprise magic and a pile of $ on smart SREs and Developers working towards a common goal. Blackboxing it will almost never work out how you want it to, at least not for the money you're probably willing to spend on it. If your RPO must be 0 then you simply have to have synchronous writes - there is no other way around it. If your RPO can be > 0 then you can do async and the world get ways easier.
|
# ? Feb 16, 2021 01:36 |
|
Sickening posted:It can be just as replicated as you could get onsite. The issue is only how big the mpls circut and how much data you are averaging. With an express route we could replicate without issue but we also weren't replicating huge, changing workloads. We could do it onsite but we're in the process of datacenter consolidation and our theory is we could use the to get around this political problem and not to mention why wouldn't we want to use the cloud? My biggest concern at the moment is if this solution in real word or is vaporware? i am a moron posted:Yes. Azure Site Recovery works fine as long as you have the bandwidth to replicate and the correct number of process & config servers for the number of servers youre protecting. Over the internet works just as well as an express route (private with private endpoints or over Microsoft peering) as long as replication is good you can have app consistency down to 15 minutes with multi VM groups. I think its 15 minutes and Im posting from my phone and this loving MS documentation is rear end in a top hat. It might be shorter than 15 minutes. I've done this with ASR in the past which honestly has worked out well but I haven't ever seen a HCI or EXSi host in Azure in the wild beyond a demo environment which is leaving me suspicious. H110Hawk posted:Drawing that green line is what separates the "$LOL" engineering orgs from the $fine ticket jockeys. You can't beat the speed of light, if you can come talk to me privately because we're about to make all the dollars. From there - what's your goal? What's your application? If you want to pay VMWare to "make it happen" they will - the question is will you choose to afford the pricetag? If not, suddenly it doesn't work and everyone is mad at you. If you set it up but refuse to do real world tests it also won't work and everyone is mad at you. You MUST pull the plug and then start DR. If you start from "do_the_needful.sh" then of course it worked - that's controlled. Go have a datacenter tech pull enough power cords on your router to shut it off, THEN run your script. I don't have a specific workload at the moment other than probably a hundreds of VMs of unknown size but you are right to the point that my requirements aren't specific enough but out of curiosity what did build with SREs? Gucci Loafers fucked around with this message at 02:50 on Feb 16, 2021 |
# ? Feb 16, 2021 02:47 |
|
Crosby B. Alfred posted:We could do it onsite but we're in the process of datacenter consolidation and our theory is we could use the to get around this political problem and not to mention why wouldn't we want to use the cloud? My biggest concern at the moment is if this solution in real word or is vaporware? Azure Replication is a real service and has been around for years. The entire service is and is marketed to be exactly what you were looking for. Again I have used it many times over the years and as long as my bandwidth was reasonable it worked just fine. It is not what I would consider vaporware at all, but if you look up the guides it does require some dedicated Systems to get the replication going. It’s a very reasonable service.
|
# ? Feb 16, 2021 03:02 |
Crosby B. Alfred posted:We could do it onsite but we're in the process of datacenter consolidation and our theory is we could use the to get around this political problem and not to mention why wouldn't we want to use the cloud? My biggest concern at the moment is if this solution in real word or is vaporware? Oh like VMWare on Azure? Gross
|
|
# ? Feb 16, 2021 03:06 |
|
i am a moron posted:Oh like VMWare on Azure? Gross You can replicate VMware virtual machines directly into replicated azure virtual machines. You don’t even have to have VMware in the cloud.
|
# ? Feb 16, 2021 03:08 |
|
What the previous poster said - this isn't about using Azure Site Recovery with VMware. I've been there, done that, it works great. What I want to know is folk's experience with VMware EXSi or whatever HCI in Azure as nested virtualization.
|
# ? Feb 16, 2021 03:09 |
|
Crosby B. Alfred posted:What the previous poster said - this isn't about using Azure Site Recovery with VMware. I've been there, done that, it works great.
|
# ? Feb 16, 2021 03:10 |
|
I know it's gross but... What is Azure VMware Solution Or... Nutanix Announces Partnership with Microsoft Azure for a Seamless Hybrid Experience
|
# ? Feb 16, 2021 03:13 |
The previous attempts at HCI/Azure have been a disaster and I work for a big VMWare/MS partner now and absolutely no one talks about this poo poo and for good reason. It’s totally pointless
|
|
# ? Feb 16, 2021 04:07 |
|
i am a moron posted:The previous attempts at HCI/Azure have been a disaster and I work for a big VMWare/MS partner now and absolutely no one talks about this poo poo and for good reason. It’s totally pointless Ah hah, that was one of the answers I was hoping to find or not find but now I know. On a related note, I remember when it first came out VMware sued Microsoft if I'm not mistaken. Any idea on the story behind that?
|
# ? Feb 16, 2021 04:12 |
|
uhhhhahhhhohahhh posted:
Fun update for this: despite two people explaining that the device doesn't become immediately unreachable the moment DHCP goes down, they just kept repeating over and over that if DHCP goes down they won't be able to get on the IP. I feel like they should know this because our old DHCP went down lot's of times and it didn't automatically stop every PC from working. So somebody went on and changed them all to statics, but hasn't documented any of the IPs anywhere.
|
# ? Feb 16, 2021 16:35 |
|
Texas DC is on generators. Everyone freak out!
|
# ? Feb 16, 2021 16:56 |
|
Woof Blitzer posted:Texas DC is on generators. Everyone freak out! Haha, half my team is in Missouri and they’re getting rolling blackouts
|
# ? Feb 16, 2021 17:22 |
|
The Fool posted:Haha, half my team is in Missouri and they’re getting rolling blackouts Weird, I haven’t had any outages.
|
# ? Feb 16, 2021 17:29 |
|
|
# ? May 29, 2024 16:09 |
|
Woof Blitzer posted:Texas DC is on generators. Everyone freak out! Well, do you need that data center? We are getting the point where generators are going to fail from the temps alone. We are getting to the points where the internet sub stations can’t stay online and residential internet , commercial internet, and cell service is starting to fail. Your Texas dc could go offline so plan accordingly.
|
# ? Feb 16, 2021 17:30 |