I have this little hobby project going on for a while now, and I thought it's worth sharing. Now at first blush this might seem like just another screenshot with neofetch.. but this thing has quite the story to tell. This laptop is no less than 17 years old.

So, a Compaq nx7010, a business laptop from 2004. It has had plenty of software and hardware mods alike. Let's start with the software.

It's running run-off-the-mill Debian 9, with a custom kernel. The reason why it's running that version of Debian is because of bugs in the network driver (ipw2200) in Debian 10, causing it to disconnect after a day or so. Less of an issue in Debian 9, and seemingly fixed by upgrading the kernel to a custom one. And the kernel is actually one of the things where you can save heaps of space when you do it yourself. The kernel package itself is 8.4MB for this one. The headers are 7.4MB. The stock kernels on the other hand (4.19 at downstream revisions 9, 10 and 13) took up a whole GB of space combined. That is how much I've been able to remove, even from headless systems. The stock kernels are incredibly bloated for what they are.

Other than that, most of the data storage is done through NFS over WiFi, which is actually faster than what is inside this laptop (a CF card which I will get to later).

Now let's talk hardware. And at age 17, you can imagine that it has seen quite a bit of maintenance there. The easiest mod is probably the flash mod. These old laptops use IDE for storage rather than SATA. Now the nice thing about IDE is that it actually lives on to this very day, in CF cards. The pinout is exactly the same. So you can use passive IDE-CF adapters and plug in a CF card. Easy!

The next thing I want to talk about is the battery. And um.. why that one is a bad idea to mod. Finding replacements for such old hardware.. good luck with that. So your other option is something called recelling, where you disassemble the battery and, well, replace the cells. The problem is that those battery packs are built like tanks and the disassembly will likely result in a broken battery housing (which you'll still need). Also the controllers inside those battery packs are either too smart or too stupid to play nicely with new cells. On that laptop at least, the new cells still had a perceived capacity of the old ones, while obviously the voltage on the cells themselves didn't change at all. The laptop thought the batteries were done for, despite still being chock full of juice. Then I tried to recalibrate them in the BIOS and fried the battery controller. Do not try to recell the battery, unless you have a spare already. The controllers and battery housings are complete and utter dogshit.

Next up is the display backlight. Originally this laptop used to use a CCFL backlight, which is a tiny tube that is driven at around 2000 volts. To its controller go either 7, 6, 4 or 3 wires, which are all related and I will get to. Signs of it dying are redshift, and eventually it going out until you close the lid and open it up again. The reason for it is that the voltage required to keep that CCFL "excited" rises over time, beyond what the controller can do.

So, 7-pin configuration is 2x VCC (12V), 2x enable (on or off), 1x adjust (analog brightness), and 2x ground. 6-pin gets rid of 1 enable line. Those are the configurations you'll find in CCFL. Then came LED lighting which required much less power to run. So the 4-pin configuration gets rid of a VCC and a ground line. And finally you have the 3-pin configuration which gets rid of the adjust line, and you can just short it to the enable line.

There are some other mods but I'm running out of characters. Why am I telling you all this? The reason is that this laptop doesn't feel any different to use than the ThinkPad x220 and IdeaPad Y700 I have on my desk (with 6c12t, 32G of RAM, ~1TB of SSDs and 2TB HDDs). A hefty setup compared to a very dated one, yet they feel the same. It can do web browsing, I can chat on Telegram with it, and I can do programming on it. So, if you're looking for a hobby project, maybe some kind of restrictions on your hardware to spark that creativity that makes code better, I can highly recommend it. I think I'm almost done with this project, and it was heaps of fun :D

During a company wide status meeting where all product managers, architects and directors assemble:

Me: *A product architect leading a team of devs*

Directors: So are there any issues or risks you see in delivering the next build in target time for Client 1?

Me: There are too many changes in feature requirements. First they said we can use a shared NFS for storage. Now they are asking to switch over to SFTP pull mode.. blah blah..

Directors: Oh I see.. well we can support both solutions then.

Me: But the deadlin..

Directors: *ignores what I say* Will be a good marketing point for future.

Me: But there are too many regressions in integra..

Directors: *ignores what I say* We should also meet deadlines. That is the most important thing.

Me: Its not as easy as 1+1=2.. The team needs more time to..

Directors: *ignores what I say* Ok lets move on to the next point. What about Client 2?

Me:

Why do devs hate windows (and all its products) so much? I mean yaa okay it is a shit os for you to get your development thingies done. Yaa I know its not open source. Yaa I know its not free. Yaa I know a lot of malwares are targetted towards windows. Yaa I know it takes decades to install updates which are released almost every week. And so on....

But wasnt windows the first operating system you laid your hands on? I mean me being a 90s born kid from India, Windows 98 was my first operating system and I was really in love with it as a teenager. MS paint was my playground and I used to go berserk over there. I mean come on. Being a teenager and knowing nothing about how a computer actually works, would you have been able to figure out how to run an NFS on linux? All the kickass presentations made in power point were so in during that time. The first code I ever wrote was in turbo C running on Windows XP. So whoever is bashing Windows and any of their products is a shit person because though Windows is not meant for developers (that too only some) it is a great operating system and I will always root for it in any forum/post where it is being bashed or criticized about. Not a Windows fan-boy (I dont known if there is even such a thing) but Windows is best for non-devs.

Attempting to access my colleague's NFS directory on his VM, don't know the VM's IP address, hostname or password:

- 2 minutes with nmap to narrow the possible IPs down to ~30

- Ping each and look for the one with a Dell MAC prefix as the rest of us have been upgraded to Lenovo. Find 2 of these, one for the host and one for the virtual machine.

- Try to SSH to each, the one accepting a connection is the Linux VM

- Attempt login as root with the default password, no dice. Decide it's a lost cause.

- Go to get a cup of tea, walk past his desk.

- PostIt note with his root password 😶

FYI this was all allowed by my manager as he had unpushed critical changes that we needed for the release that day.

Asked a guy to burn s CD with NFS underground game. puts NFS underground shortcut and burns the CD and says double click the shortcut it will install the game. 😠

Managed to make myself look like a fucking moron again today...

Can't mount NFS share, get "permission denied". Huh, that's weird... It's correctly exported.

Well it's correctly exported and rpcinfo -p $HOST times out... Must be firewall rule.

Firewall rule is changed but still no joy "permission denied"... Fuck sake networks, can't you do anything right first time?!!!

Firewall rule is correct I am reliably informed... Go about proving that it's not fucking correct and provide "evidence" to show this, I was a little bit more blunt than was strictly required.

Networks say they will take another look.

I turn NFS logging to verbose for my own interest and notice the line "path/to/directory is not a valid directory".

I, as a moron, had missed a "/" at the start of the path. That's why I still couldn't mount after the firewall change.

Go over and apologise in person and explain how I'm a total idiot.

So I found this server with a NFS share that was accessible to the public, and the good part is that the share was /root

I'm playing nowadays NFS Rivals, where I can be a racer or a cop. There's this fucking bug/feature that drives me crazy, so basically when I'm a racer, I'm a target for cops even if I'm just laying around, not racing nor crashing into things. My fucking mere existence triggers them.
Fuck you EA.

First thing I did on my first computer is to play Need For Speed Hot Pursuit. I really like NFS series since 1998, but I don't have productive PC to play the new NFS games now.

So recently I installed Windows 7 on my thiccpad to get Hyperdimension Neptunia to run (yes 50GB wasted just to run a game)... And boy did I love the experience.

ThinkPads are business hardware, remember that. And it's been booting Debian rock solid since.. pretty much forever. There are no hardware issues here. Just saying.

With that out of the way I flashed Windows 7 Ultimate on a USB stick and attempted to boot it... Oh yay, first hurdle to overcome. It can't boot in UEFI mode. Move on Debian, you too shall boot in BIOS mode now! But okay, whatever right. So I set it to BIOS mode and shuffled Debian's partitions around a bit to be left with 3 partitions where Windows could stick in one more.

Installed, it asks for activation. Now my ThinkPad comes with a Windows 7 Pro license key, so fuck it let's just use that and Windows will be able to disable the features that are only available for Ultimate users, right? How convenient would that be, to have one ISO for all the half a dozen editions that each Windows release has? And have the system just disable (or since we're in the installer anyway, not install them in the first place) features depending on what key you used? Haha no, this is Microsoft! Developers developers developers DEVELOPERS!!! Oh and Zune, if anyone remembers that clusterfuck. Crackhead Microsoft.

But okay whatever, no activation then and I'll just fetch Windows Loader from my webserver afterwards to keygen my way through. Too bad you didn't accept that key Microsoft! Wouldn't that have been nice.

So finally booted into the installed system now, and behold finally we find something nice! Apparently Windows 7 Enterprise and Ultimate offer a native NFS driver. That's awesome! That way I don't have to adjust my file server at all. Just some fuckery with registry keys to get the UID and GID correct, but I'll forgive it for that. It's not exactly "native" to Windows after all. The fact that it even has a built-in driver for it is something I found pretty neat already.

Fast-forward a few hours and it's time to Re Boot.. drivers from Lenovo that required reboots and whatnot. Fire the system back up, and low and behold the network drive doesn't mount anymore. I've read that this is apparently due to Windows (not always but often) mounting the network drive before the network comes up. Absolutely brilliant! Move out shitstaind, have you seen this beauty of an init Mr. Poet?

But fuck it we can mount that manually after every single boot.. you know, convenient like that. C O P E.

With it now manually mounted, let's watch a movie! I've recently seen Pyro's review on The Platform and I absolutely loved it. The movie itself is quite good too. Open the directory on my file server and.. oh. Windows.. you just put db.thumb on it and db.thumb:encryptable. I shit you not, with the colon and everything. I thought that file names couldn't contain colons Windows! I thought that was illegal in NTFS. Why you doing this in NFS mate? And "encryptable", am I already infected with ransomware??? If it wasn't for the fact that that could also be disabled with something as easy as a registry key, I would've thought I contracted ransomware!

Oh and sound to go with that video, let's pair up some Bluetooth headphones with that Bluetooth driver I installed earlier! Except.. haha nope. Apparently you don't get that either.

Right so let's just navigate the system in its Aero glory... Gonna need to flick the mouse for that. Except it's excruciatingly slow, even the fastest speed is slower than what I'm used to on Linux.. and it's jerky as hell (Linux doesn't have any of that at higher speed). But hey it can compensate for that! Except that slows down the mouse even more. And occasionally the mouse driver gets fucked up too. Wanna scroll on Telegram messages in a chat where you're admin? Well fuck you mate, let me select all these messages for you and auto scroll at supersonic speeds! And God forbid that you press delete with that admin access of yours. Oh maybe I'll do it for you, helpful OS I am!

And the most saddening part of it all? I'd argue that Windows 7 is the best operating system that Microsoft ever released. Yeah. That's the best they could come up with. But at least it plays le games!

Like this :/

Holy fuck EA (yea I went there)

So here I am playing NFS Heat, and my laptop is hot enough to cook eggs on the keyboard, no joke, yet spec wise I should have 0 issues running this game.

I switch on the turbo cool (basically spin the internal fans at full speed... It's an MSI) and this usually covers any heating issues, nope... not this time.

this fucker of a fan can't keep up , so I've turned down the settings from ULTRA, that's depressing, to Medium... it feels like 2006 showed back up, but here's this laptop still burning a whole in my fingers.

My GPU (GTX1060 6GB) is running at 95'c on Medium graphics, and LOW looks like I fell into a 8bit world with slow ass rendering.

Guess I'm going to need to get some cooling assistance for this thing.

Thank you EA for turning my 2.5k laptop into a stove top, I appreciate it.

I guess the name checks out, "heat"

Oh boy, linux NFS bugs! this was supposed to be fixed in 2.x, but here I am, in kernel 6.1, same issue! `nfsiod` hangs, can't SIGKILL/SIGSEGV/SIGXCPU as root, `/proc/$pid/maps` is empty, strace/gdb are denied while running as root for only this process, `/proc/$pid/stack` only shows a single kthread that's stuck waiting for a fork to return, and... what the hell is `rescuer_thread`? apparently, deprecated in kernel 3.10, whatever the hell it is... wait, how did this even make a call without memory?

it's gonna be one of those days, isn't it?

When syncing node_modules via nfs to Vagrant it's an even better excuse than 'it's compiling' - it literally takes 2-3 hrs and the ssd request is pending for over a month now. Looks like nobody sees it as an issue...

Fucking EA Games and their fucking shit mailing system!!
All the sudden they start spamming me emails about their shit games nobody fucking cares about. I proceed to inspect the footer to find an 'unsubscribe' link and there was none, just a 'manage my preferences' link.
So I went there. After waiting a whole minute for a simple page to load (wtf) there is a checkbox saying 'yes please spam my inbox with EA's latest news about their shit games nobody cares about' and it was UNCHECKED.
So I leave it unchecked and click update (thinking it might actually unsubscribe me from this crap) BUT NO! I receive another email saying 'thank you, you stupid moron you just subscribed to our shit and will now receive even more of our useless email about how different the new NFS is and how rubbish the new Star wars game is...
FUCK

a space in /etc/exports between the CIDR and the options of the NFS Server just ruined my night

I once had to fix a webservice endpoint another developer added that accepted any file from the public internet and loaded it directly onto an NFS file mount with the rest of the site's image assets and then inserted a record of the file into SQL via a hand-stitched query with parameters from the endpoint.

I was working for a large enterprise company at the time... I was very disappoint.

when you shutdown a box then remember it's NFS mounted elsewhere!

Yesterday I tried Rancher on a DigitalOcean droplet.

Rancher itself was easy to startup and use. But I couldn't manage to find the Container Volumes on my NFS Server. Even the NFS Server was up and running.

I really felt like a monkey in front of my PC not understanding what Im doing.

I hate the elasticsearch backup api.

From beginning to end it's an painful experience.

I try to explain it, but I don't think I will be able to cover it all.

The core concept is:
- repository (storage for snapshots)
- snapshots (actual backup)

The first design flaw is that every backup in an repository is incremental. ES creates an incremental filesystem tree.

Some reasons why this is a bad idea:
- deletion of (older) backups is slow, as newer backups need to be checked for integrity
- you simply have to trust ES that it does the right thing (given the bugs it has... It seems like a very bad idea TM)
- you have no possibility of verification of snapshots

Workaround... Create many repositories as each new repository forces an full backup.........

The second thing: ES scales. Many nodes / es instances form a cluster.

Usually backup APIs incorporate these in their design. ES does not.

If an index spans 12 nodes and u use an network storage, yes: a maximum of 12 nodes will open an eg NFS connection and start backuping.

It might sound not so bad with 12 nodes and one index...

But it get's pretty bad with 100s of indexes and several dozen nodes...

And there is no real limiting in ES. You can plug a few holes, but all in all, when you don't plan carefully your backups, you'll get a pretty f*cked up network congestion.

So traffic shaping must be manually added. Yay...

The last thing is the API itself.

It's a... very fragile thing.

Especially in older ES releases, the documentation is like handing you a flex instead of toilet paper for a wipe.

Documentation != API != Reality.

Especially the fault handling left me more than once speechless...

Eg:
/_snapshot/storage/backup
gives you a state PARTIAL
/_snapshot/storage/backup/_status
gives you a state SUCCESS

Why? The first one is blocking and refers to the backup status itself. The second one shouldn't be blocking and refers to the backup operation.

And yes. The backup operation state is SUCCESS, while the backup state might be PARTIAL (hence no full backup was made, there were errors).

So we have now an additional API that we query that then wraps the API of elasticsearch. With all these shiny scary workarounds like polling, since some APIs are blocking which might lead to a gateway timeout...

Gateway timeout? Yes. Since some operations can run a LONG (multiple hours) time and you don't want to have a ton of open connections hogging resources... You let the loadbalancer kill it. Most operations simply run in ES in the background, while the connection was killed.

So much joy and fun, isn't it?

Now add the latest SMR scandal and a few faulty (as in SMR instead of CMD) hdds in a hundred terabyte ZFS pool and you'll get my frustration level.

PS: The cluster has several dozen terabyte and a lot od nodes. If you have good advice, you're welcome - but please think carefully about this fact.

I might have accidentially vaporized people sending me links with solutions that don't work on large scale TM.

Fuck ssh. It does 4 things at once and i couldn't get it to do one. I have some pi's and want a shared directory on each of them. On a server i created a user for that and mounted its home directory on a pi, it worked. I did some lockdowns (no shell, only sftp allowed, login only via keyfile), but i was still able to mount it on boot.

Now i had to migrate this setup to another server. It took me a while copying all the configuration etc. All i got for that was a error-message. I figured out the users home-directory had to be owned be root, fixed that, got another error message. Somehow scp didn't use sftp but the login shell which is /usr/sbin/nologin. That made scp (and sshfs) fail, even though it perfectly works with the other server.

I gave up and removed all the setup. I'll find another distributed filesystem for that (but not samba or nfs, those are way to complicated). Those are the setbacks that depress me.

My final year taking a B.Sc. I'm writing up my Distributed Systems project, the day before handing it in. It's on top of Transis, and source code is "stored" in RCS (yes, I'm that old). The project is a reliable system administration tool, that performs the same action across a cluster with guaranteed semantics.
I'm very proud of the semantics, but cannot figure out why the subdirectory installation stuff works almost but not quite. Here's my sequence of actions:
1. Install across all machines.
2. Manually see it's broken.
3. "rm -rf *".
4. Repeat.

What in to discover is that the subdirectory installation always finishes off in a current directory 1 level higher than where it started. Oh, and the entire cluster sees my NFS home directory. Oh, and I'm running each cluster member in a deep subdirectory of my dev directory. Oh, and my RCS files live in a subdirectory of my dev directory.

All of a sudden, my 5 concurrent "rm -rf *"s were printing weird error messages about ENOENT and not being able to find some inodes. In a belated flash of brilliance, I figure out all the above, and also that I've just deleted my dev directory. 5 times, concurrently. And the RCS files.

That was the day a kindly sysadmin taught me than NetApps have these .snapshot directories.

Duck! this sloppy whiny winnfsd.

Yay! Let's use state of the art Docker with a VirtualBox VM on Windows10.

Don't get me wrong.
The Docker containers in this VM doing a great job on performance.

But in the very moment a Docker container uses a mounted folder via the windows network filesystem, all hell is breaking loose.

Building a vendor folder using a composer Docker image with 84 Packages takes about 15 seconds when cache has been warmed up.

The same Docker command pointing on a folder mounted to Windows Filesystem with warmed up cache takes about 10 Minutes!@&&@""+&

And what is the duckin' reason for this delay?
Because every transfer of a teeny tiny file has to establish a connection to fat ass Windows OS and has to pass it's glorious "security" layer.

DUCK it!
For real.

I currently working on a shell script which builds the whole vendor folder on a volume on Docker VM.
After completion, the shell script will compress the folder to one file.
This one file will be transferred over this god damned network filesystem.
Finally the script will unpack the compressed vendor folder in it's destination folder.

*sigh*
What year is it?!??

Do you consider mounting users' homedir through a nfs server a bad practice ?

Anyone has any idea how to debug occasional (Severa times a day) where one of our servers decides to mount a second copy of the same NFS? It triggers our monitoring system thinking there was a change to the mounts of the system, and I was able to verify through the mounts command, that it indeed had the same NFS mount mounted twice, with exactly the same parameters.

Is there a debug interface of some sort to see what initiated that mount? Or any tool to help me track it down? I've been stuck with this mysterious issue for a while now (As it's not really a priority, it doesn't break anything, but it bugs me and I wanna know)

Multiple projects on fire. WHY, GOD, WHY?

NFS for VCS

Anyone know of a good docker backup tool?

I'm working with fairly limited space here so Im looking for a tool to allow me to backup and restore my docker instance without storing easily recreatable data.

That means that what I want to backup are:
- Container specs (inspections or the like)
- Contents of persistent, non-nfs volumes

Since the image and the contents of my nfs volumes are stored elsewhere, and my space is limited, I decided not to back them up. Same for non-persistent (unnamed) volumes since they mostly contain stuff that you don't want to recreate on every container update or recreation but can easily be recreated if needed.

The hope is that this approach should give me a pretty slim backup while still preserving everything I need to recreate my whole docker instance.

Thanks in advance!

Just bought Intel NUC and installed CentOS on it. Mounted NFS from my NAS. Installed docker, downloaded factorio image and started server. Now I have my own game server lol.

Getting Cluster Container Servers running with a static NFS Server in the Background.

This could be called "clusterfuck", cause my mind was fucked.

Spent last hour looking for NFS Heat for PC but no sale. Saw one offer from Kastoff for $24 but seems shady...

Also seems need Origin launcher so could get banned/deactivated I think...

Then just thought why don't I just play Forza 4 that I paid $100 for a few years ago....

But yes anyone have the game, is it worth it? 4K?

Todays website fail, looks like someone is using a NFS mount for the site:
"Deprecated: Non-static method PageLinesTemplate::current_admin_post_type() should not be called statically, assuming $this from incompatible context in /nfs/c03/h03/mnt/166492/domains/makerspace.com/html/wp-content/themes/pagelines/admin/class.options.metapanel.php on line 30"

Playing NFS (it was the version which had the McClaren car), few other games and watching some movies (CD player) It wasn't my computer though. It was my cousin's and I used it while he was working. I think I broke it couple of times (windows 95) to get the BSOD.

I bought my own computer only when I started working. My family couldn't afford one before that. Luckily I had good friends in college who let me use theirs for course work.