I setup this awesome stuff with Ansible to configure (most of) my server finally. I set it up to run grafana, influxdb, and telegraf because I have been wanting to investigate those and do more stats monitoring. I got it all up and working and then realized that I have no goddamn clue how to actually make a useful graph of any of this. I can click around in grafana, I can create graphs, but I don't actually know how to say "I want a graph of Foos per Bar" and make that. Where Foos per Bar might be nginx requests per second (monitoring by hitting the nginx status module stuff) or polling for current nginx active connections or average cpu usage or free memory.
Not embarassed to say I have no idea what you're talking about.
I setup this awesome stuff with Ansible to configure (most of) my server finally. I set it up to run grafana, influxdb, and telegraf because I have been wanting to investigate those and do more stats monitoring. I got it all up and working and then realized that I have no goddamn clue how to actually make a useful graph of any of this. I can click around in grafana, I can create graphs, but I don't actually know how to say "I want a graph of Foos per Bar" and make that. Where Foos per Bar might be nginx requests per second (monitoring by hitting the nginx status module stuff) or polling for current nginx active connections or average cpu usage or free memory.
Not embarassed to say I have no idea what you're talking about.
Mmmmhmmm. Mmmmhmmm. Mmmm. Mmmm. Yeah. MMMhmmm.
I know some of these words!
I do something similar but my stack is Prometheus based which I like the query language for a lot more then InfluxDB (the exposition model is also pretty sweet).
Word of warning: free memory on Linux is not actually explained by the "FreeMemory" metric you get from VmStat.
I setup this awesome stuff with Ansible to configure (most of) my server finally. I set it up to run grafana, influxdb, and telegraf because I have been wanting to investigate those and do more stats monitoring. I got it all up and working and then realized that I have no goddamn clue how to actually make a useful graph of any of this. I can click around in grafana, I can create graphs, but I don't actually know how to say "I want a graph of Foos per Bar" and make that. Where Foos per Bar might be nginx requests per second (monitoring by hitting the nginx status module stuff) or polling for current nginx active connections or average cpu usage or free memory.
Not embarassed to say I have no idea what you're talking about.
Mmmmhmmm. Mmmmhmmm. Mmmm. Mmmm. Yeah. MMMhmmm.
I know some of these words!
I do something similar but my stack is Prometheus based which I like the query language for a lot more then InfluxDB (the exposition model is also pretty sweet).
Word of warning: free memory on Linux is not actually explained by the "FreeMemory" metric you get from VmStat.
I filter the event log for warnings and errors and then google them.
looks at the ~12 exclusively remote users we have and the other 25 who split time between being and the office and working remote
Whoops, too late.
if they come in every few weeks it's not so bad
once we get on 10 one my projects is going to be trying to get the pre-login VPN shit working
it's garbage in 7
We have a lot of users on the FortiNet client, it works really well.
Do you have an issue with it displaying an annoying pop-up over and over if the user's wifi connection is spotty, or they're on some free public wifi where they have to click something on a splash page?
looks at the ~12 exclusively remote users we have and the other 25 who split time between being and the office and working remote
Whoops, too late.
if they come in every few weeks it's not so bad
once we get on 10 one my projects is going to be trying to get the pre-login VPN shit working
it's garbage in 7
We have a lot of users on the FortiNet client, it works really well.
Do you have an issue with it displaying an annoying pop-up over and over if the user's wifi connection is spotty, or they're on some free public wifi where they have to click something on a splash page?
I've never heard about it but I'm high up the escalation chain
HELP! I don't know what the hell I am doing! This may not be the right place, but I'm going to start here as this is sys admin/dev ops type stuff.
I setup this awesome stuff with Ansible to configure (most of) my server finally. I set it up to run grafana, influxdb, and telegraf because I have been wanting to investigate those and do more stats monitoring. I got it all up and working and then realized that I have no goddamn clue how to actually make a useful graph of any of this. I can click around in grafana, I can create graphs, but I don't actually know how to say "I want a graph of Foos per Bar" and make that. Where Foos per Bar might be nginx requests per second (monitoring by hitting the nginx status module stuff) or polling for current nginx active connections or average cpu usage or free memory.
This is not a technical "I do not know where to click", this is a conceptual "I do not understand what math or functions to use to determine these things".
I realize this is as much a math question as anything, I am just hoping there is a simple rule I am forgetting (because I haven't done the math in some time now) and my google-fu is failing me on, to calculate the majority of stats I would want.
I believe I've got nginx requests/second, current active requests, and current waiting requests all correct. I pulled in a community dashboard for most server load, free memory, etc but I'd rather have the understanding to whip up my own.
Update:
Did digging through existing dashboards and I think I was just overcomplicating most stuff, probably.
At some point I plan to get to Ansible but I haven't gotten there with my lab. I would highly suggest pairing down the scope for lab stuff though. Always start small and ramp it up as you learn the technology.
Large scope is only going to lead to being overwhelmed by the setup of the tech.
Mostly just huntin' monsters.
XBL:Phenyhelm - 3DS:Phenyhelm
user got phished (surprisingly not his fault, just real bad luck*), his account started sending a million more phishing emails, all his clients blacklisted him
then when it finally gets routed to me he's in a panic and freaking out
took some effort to determine it was just simple phish and not malware on his machine, restore his access, reset all his passwords, confirmed no new spam was being sent by him
now he gets the particular joy of going to all these various places and trying to get unbanned
*He called the service desk to get a password reset for some service and they send the reset thing via email. So right when they say "ok check your mail the reset thing will be in there" the fucking phishing mail lands in his inbox. :rotate:
life's a game that you're bound to lose / like using a hammer to pound in screws
fuck up once and you break your thumb / if you're happy at all then you're god damn dumb
that's right we're on a fucked up cruise / God is dead but at least we have booze
bad things happen, no one knows why / the sun burns out and everyone dies
if you wanna get into docker, kubernetes, and ansible just start making excuses to use them and before you know it you're dockerizing the cloud and vertical scaling is coming out of every orifice
ain't never gonna happen if it's happening next week
user got phished (surprisingly not his fault, just real bad luck*), his account started sending a million more phishing emails, all his clients blacklisted him
then when it finally gets routed to me he's in a panic and freaking out
took some effort to determine it was just simple phish and not malware on his machine, restore his access, reset all his passwords, confirmed no new spam was being sent by him
now he gets the particular joy of going to all these various places and trying to get unbanned
*He called the service desk to get a password reset for some service and they send the reset thing via email. So right when they say "ok check your mail the reset thing will be in there" the fucking phishing mail lands in his inbox. :rotate:
My understanding is that is far easier to just issue that dude a new email address and IP (if you're static, if you're dynamic you might have to pull it out of rotation for awhile) than it is to get off those spam blacklists.
user got phished (surprisingly not his fault, just real bad luck*), his account started sending a million more phishing emails, all his clients blacklisted him
then when it finally gets routed to me he's in a panic and freaking out
took some effort to determine it was just simple phish and not malware on his machine, restore his access, reset all his passwords, confirmed no new spam was being sent by him
now he gets the particular joy of going to all these various places and trying to get unbanned
*He called the service desk to get a password reset for some service and they send the reset thing via email. So right when they say "ok check your mail the reset thing will be in there" the fucking phishing mail lands in his inbox. :rotate:
My understanding is that is far easier to just issue that dude a new email address and IP (if you're static, if you're dynamic you might have to pull it out of rotation for awhile) than it is to get off those spam blacklists.
Well hopefully he's only on these client internal filters and not on like, global lists for like barracuda or gmail or something. The spam was only active for a few hours.
Aioua on
life's a game that you're bound to lose / like using a hammer to pound in screws
fuck up once and you break your thumb / if you're happy at all then you're god damn dumb
that's right we're on a fucked up cruise / God is dead but at least we have booze
bad things happen, no one knows why / the sun burns out and everyone dies
+1
Options
jungleroomxIt's never too many graves, it's always not enough shovelsRegistered Userregular
if you wanna get into docker, kubernetes, and ansible just start making excuses to use them and before you know it you're dockerizing the cloud and vertical scaling is coming out of every orifice
ain't never gonna happen if it's happening next week
Yeah, docker is the new hotness in the office right now.
if you wanna get into docker, kubernetes, and ansible just start making excuses to use them and before you know it you're dockerizing the cloud and vertical scaling is coming out of every orifice
ain't never gonna happen if it's happening next week
Yeah, docker is the new hotness in the office right now.
I've been playing with it a bit.
We run Docker + Ansible for everything on staging/production.
Frontend team has a local version of the whole website + microservices running, everything in Docker via docker-compose. (They need the microservices since the frontend talks to them.)
We're primitive savages here on the backend team with shitty things like "locally installed databases" and crap, but we're slowly moving things over to docker-compose.
huh... user clicks on Excel export on AttendanceOnDemand payroll website. Website spits out an xml document instead of xls/x. Happens with multiple browsers.
He claims he plugged his old PC back in and it worked fine.
PC configuration issue with his new pc? Or he's lying about his old PC and there's something jacked up on the website.
huh... user clicks on Excel export on AttendanceOnDemand payroll website. Website spits out an xml document instead of xls/x. Happens with multiple browsers.
He claims he plugged his old PC back in and it worked fine.
PC configuration issue with his new pc? Or he's lying about his old PC and there's something jacked up on the website.
he's lying
not a doctor, not a lawyer, examples I use may not be fully researched so don't take out of context plz, don't @ me
anytime a user speaks they are either wrong or lying
often both
unrelated, this week's cisco prime problem: it is overwriting timestamps on older alarms making it appear that access points have been down for less time than they really have and confusing the operations team
So last Friday guess who had clean up someone else's mess because they "my shift is over, peace out!"?
I'm not exactly against leaving when your shift is done but, if you are gunning for a promotion you may want to show some initiative and take care of the project you're working on first.
This wound up being the 3rd time I worked over that week too. The best part was when I got in that morning I told him "hey I'm going to leave an hour early today since I've been working over all week" .
can the project not wait until the next day or something?
Unfortunately no. Due to the on-site guy waiting till the last minute to do the hardware refresh then having to bail to catch his plane, we had an a site that could not process credit cards because there was an issue with the new machines. I thought "hey lets plug the old ones back in and get them up and running" but the on-site had boxed them and UPS had already picked them up so it wasn't an option. I got them fixed but still it was a disaster.
I'm a very firm believer in leaving when my shift is done.
Agreed. However if you have applied for a promotion you need to show that you will work for it and go above and beyond. Bosses look at that pretty fondly.
can the project not wait until the next day or something?
Unfortunately no. Due to the on-site guy waiting till the last minute to do the hardware refresh then having to bail to catch his plane, we had an a site that could not process credit cards because there was an issue with the new machines. I thought "hey lets plug the old ones back in and get them up and running" but the on-site had boxed them and UPS had already picked them up so it wasn't an option. I got them fixed but still it was a disaster.
I'm a very firm believer in leaving when my shift is done.
Agreed. However if you have applied for a promotion you need to show that you will work for it and go above and beyond. Bosses look at that pretty fondly.
To which I would likely reply "poor planning on your part does not constitute an emergency on mine" and then clock out... maybe if I'm getting double time I'd stay
Is the on-site guy the one who wanted the promotion? Or is that unrelated to the person who just clocked out (catching a plane seems different than just chillaxing at shift end).
bowen on
not a doctor, not a lawyer, examples I use may not be fully researched so don't take out of context plz, don't @ me
I'll stay late sometimes to get some shit done if I think it's important.
I'm also salaried and can leave early if all my shit is done for the day and my boss doesn't care.
Respect's gotta run both ways.
life's a game that you're bound to lose / like using a hammer to pound in screws
fuck up once and you break your thumb / if you're happy at all then you're god damn dumb
that's right we're on a fucked up cruise / God is dead but at least we have booze
bad things happen, no one knows why / the sun burns out and everyone dies
Some fucking guy just had a popup that said his windows license was expired and he called the phone number and let the people on the other end of the phone remote in to his PC and gave them his credit card number to pay to fix it YOU HAVE AN IT DEPARTMENT HOW ARE YOU THIS FUCKING BAD AT LIFE AND SHIT GAWD
The only time a user is correct is when there is a confluence of both lies and wrongness. User correctness is a consequence of a user trying to lie about something they are wrong about and accidentally telling the truth.
Posts
Mmmmhmmm. Mmmmhmmm. Mmmm. Mmmm. Yeah. MMMhmmm.
I know some of these words!
I do something similar but my stack is Prometheus based which I like the query language for a lot more then InfluxDB (the exposition model is also pretty sweet).
Word of warning: free memory on Linux is not actually explained by the "FreeMemory" metric you get from VmStat.
I filter the event log for warnings and errors and then google them.
Do you have an issue with it displaying an annoying pop-up over and over if the user's wifi connection is spotty, or they're on some free public wifi where they have to click something on a splash page?
I've never heard about it but I'm high up the escalation chain
At some point I plan to get to Ansible but I haven't gotten there with my lab. I would highly suggest pairing down the scope for lab stuff though. Always start small and ramp it up as you learn the technology.
Large scope is only going to lead to being overwhelmed by the setup of the tech.
XBL:Phenyhelm - 3DS:Phenyhelm
user got phished (surprisingly not his fault, just real bad luck*), his account started sending a million more phishing emails, all his clients blacklisted him
then when it finally gets routed to me he's in a panic and freaking out
took some effort to determine it was just simple phish and not malware on his machine, restore his access, reset all his passwords, confirmed no new spam was being sent by him
now he gets the particular joy of going to all these various places and trying to get unbanned
*He called the service desk to get a password reset for some service and they send the reset thing via email. So right when they say "ok check your mail the reset thing will be in there" the fucking phishing mail lands in his inbox. :rotate:
fuck up once and you break your thumb / if you're happy at all then you're god damn dumb
that's right we're on a fucked up cruise / God is dead but at least we have booze
bad things happen, no one knows why / the sun burns out and everyone dies
ain't never gonna happen if it's happening next week
My understanding is that is far easier to just issue that dude a new email address and IP (if you're static, if you're dynamic you might have to pull it out of rotation for awhile) than it is to get off those spam blacklists.
Well hopefully he's only on these client internal filters and not on like, global lists for like barracuda or gmail or something. The spam was only active for a few hours.
fuck up once and you break your thumb / if you're happy at all then you're god damn dumb
that's right we're on a fucked up cruise / God is dead but at least we have booze
bad things happen, no one knows why / the sun burns out and everyone dies
Yeah, docker is the new hotness in the office right now.
I've been playing with it a bit.
We run Docker + Ansible for everything on staging/production.
Frontend team has a local version of the whole website + microservices running, everything in Docker via docker-compose. (They need the microservices since the frontend talks to them.)
We're primitive savages here on the backend team with shitty things like "locally installed databases" and crap, but we're slowly moving things over to docker-compose.
He claims he plugged his old PC back in and it worked fine.
PC configuration issue with his new pc? Or he's lying about his old PC and there's something jacked up on the website.
he's lying
It also gave me an XML. So something is wrong on AOD's side.
The user is lying that he tried it on his old PC.
Edit: for completeness sake just tried another browser. Same result. It's AOD's issue and not the user's.
XBL:Phenyhelm - 3DS:Phenyhelm
You guys are the best when you're not being gaping dickholes. :heartbeat:
often both
unrelated, this week's cisco prime problem: it is overwriting timestamps on older alarms making it appear that access points have been down for less time than they really have and confusing the operations team
XBL:Phenyhelm - 3DS:Phenyhelm
I'm not exactly against leaving when your shift is done but, if you are gunning for a promotion you may want to show some initiative and take care of the project you're working on first.
This wound up being the 3rd time I worked over that week too. The best part was when I got in that morning I told him "hey I'm going to leave an hour early today since I've been working over all week" .
We've done too much in our field to feed into that and now they expect it too much.
XBL:Phenyhelm - 3DS:Phenyhelm
You guys are gaping dickholes.
That is why you love us. Don't kid yourself.
XBL:Phenyhelm - 3DS:Phenyhelm
Unfortunately no. Due to the on-site guy waiting till the last minute to do the hardware refresh then having to bail to catch his plane, we had an a site that could not process credit cards because there was an issue with the new machines. I thought "hey lets plug the old ones back in and get them up and running" but the on-site had boxed them and UPS had already picked them up so it wasn't an option. I got them fixed but still it was a disaster.
Agreed. However if you have applied for a promotion you need to show that you will work for it and go above and beyond. Bosses look at that pretty fondly.
To which I would likely reply "poor planning on your part does not constitute an emergency on mine" and then clock out... maybe if I'm getting double time I'd stay
Is the on-site guy the one who wanted the promotion? Or is that unrelated to the person who just clocked out (catching a plane seems different than just chillaxing at shift end).
I'm also salaried and can leave early if all my shit is done for the day and my boss doesn't care.
Respect's gotta run both ways.
fuck up once and you break your thumb / if you're happy at all then you're god damn dumb
that's right we're on a fucked up cruise / God is dead but at least we have booze
bad things happen, no one knows why / the sun burns out and everyone dies
Welp, there goes your weekend (yes, I know it's Tuesday).
But just think, now *everyone* who uses the terminal server will have access to some kickass coupons.
Actually he's changing my oil and detailing my car today. True story.
Worst case scenario is revert to backup and everyone knows to blame that paste-eater for the work they lost.