|
Ha, slammed. It looked like for residential you can put whatever you want in your house, fume it up. It would have also been somewhat annoying to deal with if you're dealing with closed wall cavities. Nothing wrong with RG6 install though, it still can be used for satellite receivers, cable, antenna, MoCA, bringing video/media between rooms, cameras, etc. Way easier to not use it while it is in the wall than it is to install it once the wall is there any you need it. Cable jacket rating is important in commercial for sure. It seems like a lot of the one-man-crew IT guys like to just wing cable into the ceiling, zip tie it to pipeworks, etc which is just bad practice for the trade. CMP+ goes anywhere, CMR doesn't, and the CM or unrated is the cheap poo poo but not really a good idea. While my methodology sucks, I like to think I'm trying to encourage people to do better but I realize I'm not going about that the right way.
|
#
?
Nov 18, 2019 04:02
|
|
|
|
| # ? May 26, 2024 01:14 |
|
|
Avaya have an automatic deployment system for their new SIP-based handsets that works similar to the rest of the industry (Snom, Yealink, Polycom, whatever) - essentially the phone will contact a provisioning server on first boot, and if the MAC is registered by a partner to connect to a specific server, the phone will go off and connect to that. The idea is that you can ship 200 phones out and as long as you've got a record of the details off the labels, you don't need to do any configuration work on the devices themselves. So far so good. Avaya also have a 'new'  -based SMB telephone service which is a spectacular bodge of their legacy IP Office code chopped up into separate services and deployed as containers on GCP. The deployment service doesn't work with this cloud product, and the cloud product team didn't even know that the deployment service existed when they were asked about it.This company deserves to fail.
|
|
#
?
Nov 18, 2019 10:46
|
|
|
Codes vary by jurisdiction, but generally you can use general-rate (CM/CMG) cable for residential applications. But in residential multi-tenant situations like apartment buildings, riser-rate (CMR) is required. Plenum-rate (CMP) is required in plenum spaces, which is usually not a concern in residential buildings. This is what I was taught, but as always, defer to whatever authority has jurisdiction in your area. IIRC plenum-rate cable actually produces fumes that are worse for you, but burns slower, reducing the spread of flames and allowing you more time to get out of the building.
|
|
#
?
Nov 18, 2019 11:12
|
|
|
You know I don’t know if it has ever come up in one of these threads, about a building being on fire and the IT guy trying to put it out or save backups or something. All of these fire and code ratings, extinguishers, etc are meant to give you a bit more time to get out of the building not hang out or be a hero Coworker did this, saw smoke coming from the UPS room, and went in for some reason before calling the alarm. Needed treatment for inhaling smoking battery vape. Partycat fucked around with this message at 12:58 on Nov 18, 2019 |
|
#
?
Nov 18, 2019 12:41
|
|
|
But think of my file server, Hobbes, and my domain controller, Calvin! What about all the VMs running on Kirk, Spock and Bones! These are my children!
|
|
#
?
Nov 18, 2019 17:57
|
|
|
Agrikk posted:But think of my file server, Hobbes, and my domain controller, Calvin! I don’t know how exactly my mind got there but now I want to start a service that sledgehammer or shotguns your gear at random to torture people who name their servers.
|
|
#
?
Nov 18, 2019 18:02
|
|
|
Agrikk posted:But think of my file server, Hobbes, and my domain controller, Calvin! One of the hardest parts about moving to the cloud is getting clients to understand that servers no longer matter. This metal block in a rack that you've loved and nourished is now just a mouse click and a line of code. It's a really interesting concept.
|
|
#
?
Nov 18, 2019 18:11
|
|
|
poo poo pissing me off: other teams trying to assign me work at 10p without checking with me first. (I'm 7a-3p). I don't mind putting in a little overtime to flex if it needs doing, but ffs would it kill you to ASK instead of just assuming?
|
|
#
?
Nov 18, 2019 20:14
|
|
|
freeasinbeer posted:I don’t know how exactly my mind got there but now I want to start a service that sledgehammer or shotguns your gear at random to torture people who name their servers. Netflix beat ya to it.
|
|
#
?
Nov 18, 2019 20:42
|
|
|
freeasinbeer posted:I don’t know how exactly my mind got there but now I want to start a service that sledgehammer or shotguns your gear at random to torture people who name their servers. I don't mind people who name their servers, just don't make it named after gods. There's like a 95% chance on reddit that people name one of their homelab servers "zeus".
|
|
#
?
Nov 18, 2019 20:45
|
|
|
cage-free egghead posted:I don't mind people who name their servers, just don't make it named after gods. There's like a 95% chance on reddit that people name one of their homelab servers "zeus". We definitely have servers here named Thor and Valhalla 
|
|
#
?
Nov 18, 2019 20:58
|
|
|
Jerk McJerkface posted:One of the hardest parts about moving to the cloud is getting clients to understand that servers no longer matter. This metal block in a rack that you've loved and nourished is now just a mouse click and a line of code. It's a really interesting concept. I have a customer who has a senior DBA who is blocking all efforts to move to the cloud. The reason? He like to RDP into SQL servers and launch SQL Server Management Studio on the server as well as using Task Manager to check server health. Won’t look at RDS SQL Server at all because “you can’t do as much.” Yeah like worry about patching, uptime, disk space, backups, instance failures, etc. 
|
|
#
?
Nov 18, 2019 21:06
|
|
|
Man we are slowly moving all the Line of Business applications over to RDS and Mail/Office to the cloud and it is going to be glorious when it happens. My favorite part of the whole thing so far was moving a lovely estimating tool from on-prem to cloud and putting a link to the application on RDS. Then we start getting bitchy tickets about things being "unusable" and "slow" and "the server must be pinned". Nope, try again, the server is someone else's problem now, you can contact the vendor direct, gently caress off out my tickets.
|
|
#
?
Nov 18, 2019 21:23
|
|
|
Agrikk posted:I have a customer who has a senior DBA who is blocking all efforts to move to the cloud. The reason? He like to RDP into SQL servers and launch SQL Server Management Studio on the server as well as using Task Manager to check server health. RDS SQL Server is a pain in the rear end and has weird limitations. That being said his reasons were the same ones our DBA threw up. We no longer have a DBA.
|
|
#
?
Nov 18, 2019 22:14
|
|
|
Jerk McJerkface posted:One of the hardest parts about moving to the cloud is getting clients to understand that servers no longer matter. This metal block in a rack that you've loved and nourished is now just a mouse click and a line of code. It's a really interesting concept. My only real "server" experience is with AD so the idea of servers as cattle is completely foreign to me and it's absolutely holding my career back so that's cool.
|
|
#
?
Nov 18, 2019 22:18
|
|
|
Nemo2342 posted:We definitely have servers here named Thor and Valhalla Odin and Loki here, and our ESXi server that hosts them is Asgard
|
|
#
?
Nov 18, 2019 22:44
|
|
|
Agrikk posted:I have a customer who has a senior DBA who is blocking all efforts to move to the cloud. The reason? He like to RDP into SQL servers and launch SQL Server Management Studio on the server as well as using Task Manager to check server health. Last year a I had to install an application on a RHEL server, but weirdly enough it required xwindows to install it. It was an oracle package of some sort. It stopped a project short because there's no GUI on AWS servers, but I just setup X11 fowarding and took care of it. It was odd, I couldn't even imagine designing a linux app like that.
|
|
#
?
Nov 18, 2019 23:03
|
|
|
cage-free egghead posted:I don't mind people who name their servers, just don't make it named after gods. There's like a 95% chance on reddit that people name one of their homelab servers "zeus". Just lol if this isn't the base for all of your server naming needs: https://en.wikipedia.org/wiki/List_of_Greek_mythological_figures#Immortals I use Ares, Erebos, and Hemera I don't care about naming on my private servers. Work is config + dc + fqdn, e.g. centos-xl-fra1-1-db2.your.name.here
|
|
#
?
Nov 18, 2019 23:34
|
|
|
My kingdom for a sane way to sync GAL's between O365 tenants.
|
|
#
?
Nov 19, 2019 01:17
|
|
|
freeasinbeer posted:RDS SQL Server is a pain in the rear end and has weird limitations. That being said his reasons were the same ones our DBA threw up. The essence of RDS SQL server is that it behaves as a mirrored cluster and anything requiring local admin or access to the OS has been obfuscated. Instead, all the things you probably want to adjust are in your options group, cluster group or are available as stored procedures that manage all the nuances of operating in a mirrored environment for you. But the only things you really need to do is get your disk size right (so that you aren’t starved for iops) and create an option group with S3 permissions so you can schedule nightly backups straight to S3. And you aren’t using SQL Server Agent anymore anyways, right? My favorite use case is from a customer of mine, who programmatically bumps up the database class on the first of the month for an inventory database to handle new inventory data ingest and processing, and related activities. Then on the fifth the database downshifts to a maintenance level for the rest of the month’s ad-how queries. They save thousands a month over their on-Prem setup simply from recognizing the cyclical nature of their workloads. Agrikk fucked around with this message at 01:28 on Nov 19, 2019 |
|
#
?
Nov 19, 2019 01:23
|
|
|
Goddamnit Java ! I've got the installer and the .cfg file on a network share. I run a PowerShell script which runs the .exe with the path to the .cfg file. Sounds simple, right ? Having the path to the .cfg file be a UNC path is not supported, you get an error popup. No matter that it's on the same goddamned share as the .exe. Worse, it worked as recently as u202. Worse, the bug is resolved, "Unable to reproduce". Which is bullshit. Anyone with an openjdk.java.net account want to re-open this ? https://bugs.openjdk.java.net/browse/JDK-8221968
|
|
#
?
Nov 19, 2019 01:32
|
|
|
Agrikk posted:The essence of RDS SQL server is that it behaves as a mirrored cluster and anything requiring local admin or access to the OS has been obfuscated. We want to replicate across regions for reasons. That’s kind of a pain rear end as it stands.
|
|
#
?
Nov 19, 2019 01:57
|
|
|
freeasinbeer posted:We want to replicate across regions for reasons. That’s kind of a pain rear end as it stands. A huge pain in the rear end, I agree. Have you looked into replicating using the always-on functionality of DMS? I’m actually about to start a project using RDS SQL and DMS to set up a replication mesh in three regions. It might be horrible, might be great. Who knows, but where multi-region replication comes in I’m leaning towards horrible, Especially with a couple of seven billion record tables...
|
|
#
?
Nov 19, 2019 02:11
|
|
|
bobmarleysghost posted:This triggered me. Our MES frontend is Access and I have dreams of sometimes converting it into a python web app, and then I wake up in a cold sweat. Have you thought about switching to Igniton? Its great for MES, and you can dump the data into most SQL Dbs.
|
|
#
?
Nov 19, 2019 06:58
|
|
|
Speaking of SQL Server, an instance using all available RAM (0 bytes free) and having sluggish performance in RDP are not at all indicators of trouble, right? Those just seem like indicators that the instance is using all available resources to serve requests, especially when regular query performance from remote applications is fast.
|
|
#
?
Nov 19, 2019 10:00
|
|
|
Min'es called Patrica. I blame it on poo poo going wrong.
|
|
#
?
Nov 19, 2019 12:14
|
|
|
nielsm posted:Speaking of SQL Server, an instance using all available RAM (0 bytes free) and having sluggish performance in RDP are not at all indicators of trouble, right? Those just seem like indicators that the instance is using all available resources to serve requests, especially when regular query performance from remote applications is fast. Sounds like someone didn't set an upper memory bound in SQL and are possibly running it with locked pages.
|
|
#
?
Nov 19, 2019 14:53
|
|
|
http://fmsinc.com/MicrosoftAccess/Errors/query_is_corrupt/ ha Ha hahahahaha Critical Alert: Query is Corrupt, Error 3340 Microsoft's ResponseA set of Microsoft Office security updates released on November 12, 2019 causes Access databases to fail when it runs Update Queries to modify data. We are in touch with the Microsoft Access development team and do not have a clear answer on what went wrong and how something so serious was released to the world. We share your concern that security updates like this cause a bigger problem than the viruses or malware they are attempting to block.
|
|
#
?
Nov 19, 2019 15:28
|
|
|
Coworker had to figure that out yesterday. Worst part is MS was just like "Here's a workaround for running affected queries, we'll release a fix next Patch Tuesday"
|
|
#
?
Nov 19, 2019 15:35
|
|
|
nielsm posted:Speaking of SQL Server, an instance using all available RAM (0 bytes free) and having sluggish performance in RDP are not at all indicators of trouble, right? Those just seem like indicators that the instance is using all available resources to serve requests, especially when regular query performance from remote applications is fast. By default SQL Server will grab all the memory it can out of an instance for its operation. Normally it will leave a small percentage free for the OS but in certain situations it will take literally everything and starve the OS of memory. That’s what sounds like is happening here. Like bull3964 said, You can prevent that from happening by setting an upper bound for what SQL Server will take to keep the OS healthy and the SQL Server happy. I think the rule of thumb was leave 2-3 GB free for the OS on a dedicated database server. Just to be sure you probably want to check memory pressure of your database instance to see if your box could do with more ram.
|
|
#
?
Nov 19, 2019 15:45
|
|
|
Agrikk posted:A huge pain in the rear end, I agree. I forget the exact reason, but we couldn’t use DMS for some reason, and the AWS recommended option was: https://aws.amazon.com/marketplace/seller-profile?id=9336c2be-e8bf-4d8e-869c-c11d1fd975ce Which is a trash heap.
|
|
#
?
Nov 19, 2019 16:50
|
|
|
Agrikk posted:By default SQL Server will grab all the memory it can out of an instance for its operation. Normally it will leave a small percentage free for the OS but in certain situations it will take literally everything and starve the OS of memory. That’s what sounds like is happening here. Fun fact, out of the box the upper limit on SQL server is 2TB. This is why it grabs all the memory beyond what Windows has locked up. Coincidentally this is also the maximum amount of memory SQL Server can handle.
|
|
#
?
Nov 19, 2019 17:22
|
|
|
Agrikk posted:By default SQL Server will grab all the memory it can out of an instance for its operation. Normally it will leave a small percentage free for the OS but in certain situations it will take literally everything and starve the OS of memory. We had this happen.. Our EDI software company didn't bother to set a limit when they installed SQL Server. It happily gobbled up every single bit of the 48GB available causing other operations to fail.. Kinda dumb the default setting is to starve the OS.
|
|
#
?
Nov 19, 2019 17:32
|
|
|
Seriously. How hard would it have been to put in a line like Set sql_max_mem = server_mem - 4GB as a starting place? I know that you’d have to reset it if you ever added memory to the box, but if you added memory, wouldn’t you have memory on the brain and think about performance tuning your server anyway?
|
|
#
?
Nov 19, 2019 17:50
|
|
|
Agrikk posted:Seriously. How hard would it have been to put in a line like I am 99% certain that microsoft has suggested this as a best practice for years.
|
|
#
?
Nov 19, 2019 17:52
|
|
|
Sickening posted:I am 99% certain that microsoft has suggested this as a best practice for years. My point is why couldn’t Microsoft adopt this during the installation process? Why leave this in the hands of the installer as part of the post installation process? I mean, we do this for RDS SQL Server for you (in addition to a bunch of other configuration tweaks and performance enhancements) so why can’t Microsoft?
|
|
#
?
Nov 19, 2019 17:53
|
|
|
Agrikk posted:My point is why couldn’t Microsoft adopt this during the installation process? Why leave this in the hands of the installer as part of the post installation process? I assume its something to do with making the software easy to install. Its like vmware not making NTP a required opt out part of the installation.
|
|
#
?
Nov 19, 2019 17:55
|
|
|
The company I'm working for is being acquired. New CEO is doing a Q&A. When asked about remote workers, he said the company prefers everyone to be together, and only 10% of their workforce is remote. That means I'm hosed, as I'm full-time remote and there's no office within 500 miles of me. I guess this means I'm in the market for a new job. Anyone need a guy whose expertise is data analysis to find opportunities to improve products and process?
|
|
#
?
Nov 19, 2019 18:17
|
|
|
There are also different ways to configure memory and that setting doesn't mean what people think it means. The setting is for the buffer pool, memory consumed by the engine for running queries and plan caches are outside that Max setting. Also, that means that SQL is not limited to only 2tb of RAM, only the buffer pool is. The sql engine can scale to the OS max. Although it's not recommended, you probably won't notice much performance degradation (depending on workload) if you don't set the max as long as you don't have locked pages set. The OS will page out parts of the buffer pool if there are other memory demands and as long as you aren't utilizing the whole buffer pool, you probably won't get bitten. If you lock pages in memory though and don't set a max (and a host reservation if you are virtual) you are in for a world of fun things breaking in interesting ways as memory pressure mounts.
|
|
#
?
Nov 19, 2019 18:18
|
|
|
|
| # ? May 26, 2024 01:14 |
|
|
ConfusedUs posted:The company I'm working for is being acquired. Goondolences.
|
|
#
?
Nov 19, 2019 18:20
|
|