Just two days ago, I mentioned in this blog the Windows port Microsoft made for the ARM architecture:
Microsoft, for things like the (abandoned) Windows RT and Windows Phone, besides porting some of the upper layers of the Windows stack and developing new ones, also had to do additional work to get the NT kernel to run on such hardware. It’s worth mentioning that despite that effort, Windows Phone 8+ has hardware requirements higher than those of Android (comparing versions released in the same time span, please correct me if I’m wrong).
Today, as I open the web browser I’m greeted by multiple related news: a quad-core ARMv7 / 1 GB RAM version of the popular Raspberry Pi board, named “Raspberry Pi 2”, was released, will run Ubuntu Snappy Core, and, mind you, Windows 10.
Now that Windows RT is pretty much dead in the water, it looks like Microsoft found at least one use for their port besides Windows Phone: a strategically introduced “Windows 10 for makers”, which is free – something that would come out as impossible some years ago, in the license-angry Microsoft phase. (Yes, just like they also offered the Windows 8.1 license to OEMs of tablets with screen >= 7 inches, and apparently will offer Windows 10 to Windows 7 and 8.1 users for one year after its release). Of course, there’s no news of this Windows version being the slightest open-source – that is something that in the beginning of 2015, is still thought as “impossible” – but Microsoft is making promising steps, after having open-sourced the .NET framework.
Now, let me explain: while this is a nice move from Microsoft, is not something that leaves me particularly happy (in fact, it leaves me somewhat worried, and it’s not because “OMG OMG Linux is going to lose market share”). For starters, there’s the fact that there have been way more powerful ARMv7 devices around for a long time, for a similar or equal price ($35 USD) – take a look, for example, at the ODROID-C1, so why didn’t Microsoft decide to offer Windows for those too?
The answer, in my opinion, is very simple: Microsoft wants to “look cool”, and benefit from the free advertising and consequent increase of popularity a partnership with the Raspberry Pi Foundation has. Releasing Windows 10 for ARM in a more flexible setup (one that would support different boards besides the new Raspberry Pi) would be even more interesting to the community and probably more flexible for wearable, IoT, etc. projects, but that’s not the path they chose.
Supporting only the Raspberry Pi is also the easiest option, partly due to the lack of standards in the ARM world, which I complained about in the aforementioned blog post. Supporting other boards would lead to a lot of work supporting the different SoC, different peripherals, different boot methods and imaging formats (nothing that Microsoft couldn’t abstract away with a generic second-stage bootloader for WIM files), etc. In other words, it would leave them with as much work as the Linux community has in order to support different embedded systems and CPU architectures (heh).
Something that’s still unclear to me is the licensing part. This Windows version, while free, certainly comes with caveats. I’m sure Microsoft won’t allow using it on consumer products based on the Raspberry Pi (for example, using the upcoming version 2 of the Compute Module), as otherwise it would constitute a free alternative to licensing Windows Embedded. I also expect this version to come severely crippled as not to be able to act as a server; otherwise, expect some cheap Windows servers coming up soon. Even if it’s not crippled, the EULA rules it all, which means that even if a port of this Windows to other ARM boards and devices was possible, or if someone starts selling RPi-based Windows ARM servers, it most likely would not have Microsoft’s blessing.
I’m also a bit worried that the Raspberry Pi Foundation might start to push for Windows instead of Linux, especially since I bet the Windows port will be much more popular than the Linux distros. Doing so would kill half of the purpose of the whole Raspberry Pi thing, in my opinion. Most kids, given the opportunity to use the same user interface and some of the programs they are already used to, will never try to learn new things, much less tinker with them. Having them use a different GUI, or perhaps no GUI at all, by booting directly to a shell (try that, Windows! The closer you can get is the command line on a Windows Recovery Environment! heh) allows for an experience that is, from the start, much different from what they would get with a typical off-the-shelf computer.
By making people move out of their “comfort zone”, the Raspberry Pi Linux distros encouraged people to learn their way around a different system. I’m afraid people who buy a Raspberry Pi and promptly install Windows on it will keep without knowing what a command line is, and will keep doing the same things they did on their full computers. Text-based UIs are definitely not the best option for many/most things, but there are many things they’re better at than GUIs, and more importantly, some people find them out to discover they like them much more than GUIs. But if these people are never given the opportunity, they will never find it out.
Now I must admit, it would be super awesome if Microsoft came up with something like FX!32 but for running x86 binaries on ARM. That would probably require an even more restrictive EULA and/or more crippling for this Raspberry Pi version, as then people would be able to run, for example, all sorts of existing server software that currently requires a paid Windows Server license.
To conclude, I don’t think Microsoft is actually interested in making Windows available for more devices or actually making it a viable choice for low-cost embedded hobby projects and consumer products. They are just trying to gain popularity among not only the general public, but also among the developer and hobbyist community. Unfortunately, I’m not sure if this movement of “embracing OSS and open-sourcing ALL THE THINGS!” is going to last when/if Microsoft reaches their market share goals.
We can look into how the competition did: Android, initially pretty much completely open source, was made more closed as its market share increased. Google keeps on moving more and more things to closed-source blobs under their control, which has upsides (it’s easier to update many parts of the OS) and many downsides (lower user control, etc.). I wonder and worry if Microsoft will do something similar as their popularity endeavors are successful, turning their back on users, developers and all this “freedom hype” once again.
On 24th June last year, Version 1.4 of my Utilities add-in for the Casio Prizm calculators was released. The plan was for this to be final release of said software, with any further versions being bug-fixing only, and because of this, it was even more thoroughly tested than previous stable releases.
Ironically enough, an apparently innocent code optimization, introduced at a late development stage, introduced a bug in the Tasks functionality of the add-in, where a reference to a nonexistent memory object may happen when there are no tasks. At this point, I was more or less tired of the Casio Prizm platform, because of the many issues I have described throughout the years, and which the homebrew development community is yet to fully solve. However, as time went by, occasionally I’d look into my Prizm projects and I’d inevitably end up optimizing yet another function, or adding another small functionality.
This, plus the desire to iron out some edges, led to the discovery of another bug, this time in the calendar search function. After lengthy debugging sessions it turned out to be a buffer overflow issue that could happen when reading malformed calendar database entries. Fixes for these and other bugs, plus the functionality I added as I had time and will, made it clear that releasing a new version of Utilities was imperative. I often ask myself if continuing the development of such project is still worth it, since:
- I use my Prizm much, much less than I used to (finishing high school marked the end of the period in my life where graphic calculators were needed for education);
- The community of users of these calculators was never very big, and keeps on shrinking. Of the people who go on online communities dedicated to these calculators, some lost interest on the device, and others lost the device to a brick, for which nobody is able to pinpoint a certain cause. Taking into account the results of my survey so far, the intersection between the group of people who own a Prizm and the group of people who search for software for it, seems to be contain no more than 50 people;
- Of the people who remain in the communities, most never paid much attention to Utilities (due to feature creep, it’s likely that most people never understood its power) and the amount of users that still pay attention has reduced too (as well as their attention span for it).
Apparently, at least one hundred thousand of these devices are produced every month, but the amount of users who know they can run extra software in them, is in the order of the few dozens.
Despite all this, such questions are promptly answered by the fact that I still have fun developing it, even if nobody gets to use my work. And so development progresses, albeit at a much more relaxed rhythm, firstly because v1.4 is still very stable (at least, no one complained), and secondly because there is no roadmap to v1.5 nor planned release date. Heck, if I wanted to, I could not release it, and zero people would complain… but perhaps not after seeing what’s coming.
On the video below (without sound), I show a small subset of the new functionality for v1.5 (if it ever gets released, heh heh). The part that, in my opinion, is going to leave the mouths of some people open, starts at 3:30. It is an elaborate method to allow people to extend Utilities to a certain point, by having an easy way to use the big amount of utility functions used internally, as well as the nice GUI methods I developed. As if this wasn’t enough, one still gets access to most known syscalls (those that involve function pointers being the notable omission). What’s presented is, after all, the most powerful scripting engine ever made to run on the Prizm, and because of this one gets goodies like on-calculator development.
As hinted in the video, “PicoC script execution available on select builds only”. Starting with version 1.5 of Utilities, there will be two public builds made available: the normal one, with the now usual feature set plus the added features but without PicoC, and another with all that plus PicoC support enabled. The reason for this, is that such support increases the size of the add-in by at least 60 KiB, and as can be seen in the video above, the scripts have (almost) full reign on the machine, including read/write access to the whole address space (in the video, you can see a script changing the function key color, and while it’s not depicted, it also locks and unlocks Main Menu access). This means that a script can definitely brick a calculator on purpose, and do all the sorts of nasty (and good) things an add-in can do, except use syscalls with function pointers (the reason being, that PicoC doesn’t support them). It’s understandable that not everyone wants to have such a thing installed on the calculator, hence the limited builds.
PicoC is not especially fast, but definitely fast enough for many applications. It is also riddled with bugs, and even things as simple as the scope of variables appear to have bugs. Adding the differences between PicoC and the C90 standard it aims to run, expecting to write C code with the same kind of ease (if it was ever easy, especially after using newer C standards or C++) as when using a fully featured compiler is certainly unrealistic. Still, I hope my PicoC port will constitute an interesting alternative to the never-finished LuaZM and to the Casio BASIC interpreter that comes with the OS.
Regarding the other changes seen on the video, there’s the rearrangement of menus on the home screen. The tools menu now hosts a balance manager, with support for multiple wallets, and it will also host a password generator. The old tools menu has been moved to the “Memory & System” menu on the F5 key.
That’s nice and all, but for when?
I don’t have an answer to that. With v1.5 I would like to include even more features than what I have added so far, namely a proper text editor. Such an editor is being developed by ProgrammerNerd / ComputerNerd, who, just like me, doesn’t always have much free time to work on such things. So I’m patiently waiting, and you should too. Meanwhile, feel free to ask any questions, request features (please be reasonable, and I don’t promise anything) or request development builds for a sneak peek.
During the past year, the WebView vulnerability(ies) in Android have been making the rounds in various technology-focused websites. More recently, another WebView vulnerability was discovered, affecting versions 4.3 and below of the popular mobile OS (or roughly 60% of the users). Three days ago, HotHardware released a piece on why Google will not patch this vulnerability on 4.3, let alone older versions.
As a quick reminder, Android 4.3, the last version of the Jelly Bean series of releases was launched on July 24th 2013 and its last point release (4.3.1) on October that year. That was 15 months ago. A device that shipped with this Android version was the second-generation Nexus 7, which is still under warranty on places where two-year warranty is mandatory, like in the EU. The Nexus 7, being a flagship Android device from Google, received updates to more recent Android versions; the same can’t be said about most other devices released with 4.3 or earlier.
Those 60% sure would like to be in the 39%.
Most of the discussion so far has been centered around whether the responsibility to patch older Android versions and/or push new ones to phones is on Google’s side or on the manufacturers’ side, or if the problem really is with the carriers, which won’t update their customized builds of the OS. There’s also the line of discussion that says such responsibility does not exist, because the problem is fixed in the latest Android version, and anyway, For God’s sake, are you still using a phone that came out six months ago? So vintage. Oh wait, how are you not using a high-end phone from <insert major brand>? (and even high-end phones sometimes don’t get updates past the next major release)
I would like to shine light on another side of the problem: the fact that smartphones, tablets and devices alike can’t be updated by the user software-wise. In fact, it’s not just the user who can’t update or choose to run a different operating system: I’m convinced that for the most part, if the manufacturers wanted to update their Android systems to a more recent OS version, or switch to, say, Windows Phone or Firefox OS, they would have much trouble themselves. And I pinpoint this down to two different but related issues, the lack of a proper drivers system on Android (possibly involving Linux) and the multitude of ways these devices boot their OS, expect updates and do basic hardware communication. Both issues are related to a bigger problem: the lack of standards in the world of embedded consumer electronics.
In this text I’m letting aside all the arguments regarding “open source vs. closed source”, “walled garden vs. open garden”, “but but binary blobs!”, etc. Both theory and practice have evidence that these debacles and inconveniences don’t matter, or there are ways to work around them that are successfully used in practice. The only “inconvenience” that might remain, is the hardware manufacturers’ wish for people to replace their “old” devices every six months or so. This turns out to be a game of extortion made for those who worry about their security: “if you want a OS patched against this horrible vulnerability, just buy a new device that won’t do much more than your current one, but will have that single line of code changed”.
In a perfect world though, manufacturers which wanted to play that game would have to do it in the clear, by explicitly locking their devices (as most already do) and announcing on the box that there will be no updates, fixes or warranties software-wise. (Curiously, the texts that say such things are usually free-as-in-beer software licenses, not software you pay for in the form of hardware). But letting aside the utopia and focusing on the two standards-related issues I mentioned before.
I said Android doesn’t have a proper drivers system. This statement can be taken as incorrect, because, after all, Linux is the part of the stack responsible for driving the hardware. But while Linux is not Android, Android definitely includes Linux, and their creators and maintainers make a deliberate choice to use this kernel. I’m not saying it’s a bad choice, well on the contrary – only Linux and a few other Unix-like kernels could scale down and adapt to the hardware and ARM architecture used in most handheld consumer devices.
Using Linux is taking a giant shortcut (again, that isn’t bad – reusing is good). Microsoft, for things like the (abandoned) Windows RT and Windows Phone, besides porting some of the upper layers of the Windows stack and developing new ones, also had to do additional work to get the NT kernel to run on such hardware. It’s worth mentioning that despite that effort, Windows Phone 8+ has hardware requirements higher than those of Android (comparing versions released in the same time span, please correct me if I’m wrong).
Going back to the drivers, many people say the big roadblock to making new Android releases run on (relatively) old hardware is the binary blobs, the closed-source drivers that control much of the hardware in those embedded systems. Now, a bit of anecdotal evidence: I use proprietary drivers from at least Nvidia and Broadcom on the Linux install on my laptop, and these have survived fine upgrades from Linux Mint 15 to 17, and multiple Linux kernel updates from at least 3.8.8 to 3.14.27. This is because the proprietary part is well separated from the things that can possibly change between kernel versions, and there are clear update paths defined.
Of course it helps if the maker of the proprietary drivers is interested in having their drivers run in newer operating system versions, but if all drivers were properly developed and not added into the system as ugly kernel patches (or should I say, “hacks”?) for which nobody has the source, as I’ve seen System-on-Chip manufacturers do (looking at you, Mediatek, Realtek, …), the problems would be mostly gone. The practice of doing such ugly source editing is one of the reasons I say that even if manufacturers wanted to, they couldn’t switch to another OS or update to more recent Android versions. I suspect that at some companies, just a few months after devices ship, even high-end ones, entire source trees, complete git repos, are rm -r-ed out of every system. Nowhere does the GNU GPL say that it’s not a violation of the license if you get rid of the source, does it? As if such license was ever read by said people…
There is another “entertainment” awaiting those who take the updating matter into their own hands and attempt to port the OS of their liking to their device, which is understanding how the device expects to be updated and how it starts its OS. While this is sometimes just a case of watching updater software do its job (that is, when an update is even available), often additional steps are needed, and this is where one finds out that most devices use U-Boot, but often it’s even more patched than the Linux kernel, and again, source code is nowhere to be seen. There is then a myriad of ways to boot the kernel and from there to starting userspace, and fortunately this is more or less constant between Android devices. Still, undocumented quirks are everywhere, and one basically has to work with each device on an individual basis. The same model has various versions? Great, expect to repeat that work for each version.
These all have a color screen, a speaker, a microphone, some buttons, and can make calls. It’s 2015, standards exist, they must be really similar, right? Yes, as long as you don’t attempt to change their OS…
And finally, we get to what I personally think is the core of the issue: each device is too much of an individual situation, and work must be done for each device. It’s been like this since, well, ever – for well more than a decade, since what can be called the first smartphone was launched (HTC Wallaby). In the beginning, I think this was justified – the hardware was not very powerful to be able to handle complex software abstractions and advanced boot methods, nor did software advance at today’s pace. Consumer handhelds were also not as ubiquitous as today. We can compare this to the evolution of the Personal Computer, where in the end everyone settled around the IBM PC standard. A corresponding standard for the smartphones and tablets everyone has is yet to be found – such a standard is what enables one to buy almost any computer off the shelf and install a different OS in it, or a different version of the same OS. It would also allow for buying devices without OS preloaded. This means the user would be able to control its user experience and security. I would no longer have to buy a new phone to stay safe, just because (and this would happen inevitably – no software is bug-free) a vulnerability was found in Android 4.2.
Sure, despite the PC standards, there are computers in the market which come as locked down as today’s tablets and smartphones. And there is no problem with that, as long as such locked-down things are not the only option. When locked-down is the only option, or unlocked options are prohibitively expensive, there is little room for innovation, consumers end up not having much to choose from, and eventually, no way to have durable hardware, if all the available alternatives support the “update the hardware to update the software” scheme.
Even in today’s context, there are better ways to ensure operating systems keep up-to-date in terms of security, without exactly requiring a change to another version. Google should look a bit more into Microsoft, which got one thing right on Windows for over ten years: Windows Update. Microsoft ensures support for a specified number of years for its OS, independently of the hardware it runs on; this is something consumers like and enterprises love. Google seems to have learned, so much that it is moving a lot of things that were previously built into Android to Google Play Services, a component that can be updated through the Play Store like other apps. Unfortunately, this means making more and more of the OS closed-source, but that’s another subject. Personally, I would rather pay, say, 10 to 20% of the original price of my phone with each update, than having to buy a new phone when I definitely don’t need one except for the bits executing in its CPU which all of a sudden are “old” and insecure.
I believe an update scheme a-la-Microsoft would be profitable for Google and let them have a bigger market share in the enterprise. (Actually, if Google is taking any of that market share, is because of the “cloud! factor” and because enterprises are moving to Google’s systems as “it’s what everyone uses”, and not because it fits their needs better). It could be perceived as terrible for hardware manufacturers, because there would be one less reason to buy new devices, and let’s not forget Google also sells hardware. Apple sells hardware too, and people happily run Windows, Linux or whatever on their Macs and MacBooks, and I doubt Apple has lost any business because of that: well on the contrary. It shows the two things don’t need to be exclusive. Apple still manages to sell a lot of Macs and people who want to stay with an older machine still enjoy updates for much longer. In their line of consumer handhelds, while it is perceived as being even more locked down than the competition, each model tends to get at least two major OS updates (for free!), making people who aren’t in an “upgrade cycle” happier.
I am actually surprised and annoyed that consumer rights associations don’t complain more about the situation. It seems that certain companies were successful in sinking into people’s minds the idea that in the case of phones, tablets, smart watches, etc. the software can’t be decoupled from the hardware. In fact, in its current state, it’s really hard to decouple it, but it’s because that’s what manufacturers want, not because of technical obstacles. Perhaps this thinking comes from the fact that, after all, the introduction of smartphones and tablets to the general public was done by Apple, which presented their vertically-integrated walled-garden first and foremost, and giving everyone else the idea that was the only way these devices would ever be successful.
To finish, another anecdote. I have bought a cheap unknown-brand tablet with a x86-64 Intel CPU. It runs full Windows 8.1 and is fully up-to-date thanks to Windows Update; I’m very happy with it. When Windows 10 comes out I plan to install it; either the upgrade is as easy as from 8 to 8.1, or I’ll install it manually by connecting a USB stick and using the UEFI. As we know, Windows is closed-source, and drivers are nothing more than closed-source “binary blobs”. Still, I know I’ll be able to install most if not all of these drivers in Windows 10, to a point where I can use that version of Windows on the hardware I have now. Perhaps I’ll need to throw some money at Microsoft to have Windows 10, if that idea of giving it for free to users of 8.1 and 7 turns out to not apply to me. Had I bought an Android tablet, I could throw money at Google and at the manufacturer, and I’m sure that after a year or so, neither would put a single update out for the hardware. The money would have rendered a new piece of hardware, yes… but of how much use is another piece of plastic and silicon, when the previous one works perfectly? They sure like to contribute to e-waste.
Related question: are there any phones running full x86 Windows? Perhaps once Windows 10 comes out?
My previous post on this blog was published by the end of the long-gone month of June. Many things have changed since then, for example, I entered university and was pressed into creating a Facebook account (more or less separate from the rest of my online presence, so don’t look for me, I won’t add you). On that post, I rambled about the recovery from a big server outage that costed 42 hours of tny.im downtime, and over one week of server downtime. I learned my lessons (I doubt BlueVM learned theirs, but that’s a whole other story), and I went forward with what I said I would do: “setting up a new advanced and redundant system” for ensuring tny.im is always up.
That system has been up and running for over two months now, with varying amounts of servers making the redundancy and load balancing, and a plethora of occasional hiccups. Right now it’s composed of three virtual servers (all from different providers…), but there were times when it was composed of five servers. These three servers are paid, and while they aren’t exactly expensive (but not the cheapest, either), you can imagine the bill, so let’s not talk about tny.im profitability now, OK? (I have kind of given up).
In the spirit of the great statisticians of our time, here’s a graph without title, labels or axis.
However, having three servers serving the same website, with all three of them being almost a clone of each other (which means, all have the same files and database contents, synced), in a DNS round-robin setup doesn’t directly lead to greater uptime. In fact, I have found out it can lead to more outages, since now the total downtime is approximately the sum of the downtime of each server. Of course, most of these outages are partial (as in, only users unlucky enough to have their DNS request resolved to the IP of a server that is down, will actually perceive the site as down), except for when the MariaDB replication freaks out and basically grinds all database operations, on all servers, to a halt, requiring a complicated manual restart of all MariaDB instances, in a specific order (yes, I have spent many hours searching for an alternative database system, and couldn’t find any that met my requirements).
In order to actually achieve greater uptime, one must have a system that automatically manages the DNS records so that the domain(s) of the website in question never have any records pointing to servers that are down. In other words, the “sheep” must be “hidden from sight” as soon as they go “bad”, and should be put back “in stage” once they become “good”. Being DNS something that was definitely not made for real-time record edits, with many systems caching DNS request results well beyond the specified TTL, this system obviously doesn’t ensure that the “bad sheep” are not invisible to everyone watching the show. But if it manages to do it for even a small percentage of the public, it’s already better than not hiding from anyone (and especially, if it successfully hides the problem from the uptime monitor, that’s even better 🙂 ). This explains why the DNS records for tny.im are set with TTLs of five minutes.
The development of such a DNS record management system was also more or less contemplated in my previous post, when I say:
I’ll take this downtime and new server acquisition as the motivation for setting up a new advanced and redundant system, so that if one server goes down, tny.im (and possibly this blog too) will continue to operate as normal.
And in the end, in a later edit:
On related news, Mirasm – the Tiny Server Redundancy Manager – is mostly finished, only needs some more testing to be put on production servers, managing the new tny.im redundancy system.
“Mostly finished”, as we all know, really means “It’s 99% ready, I only need to figure out the remaining 1% that consists on… everything that is tricky and I’m not sure how it’s done”. This is specially true in this case, as I had high requirements for my manager: it couldn’t use any resources other than the servers I had already (it would’ve been easy to have a separate server just for monitoring and editing DNS as needed, but I didn’t want to pay for yet another server on yet another provider), and it couldn’t fail more than tny.im itself. In fact, the time when the manager has to do more important work, is when it is not working, i.e. when a server goes down and so goes the manager. I finally finished the project, and it works as planned. I only got the name wrong…
Introducing mersit, a Tiny Server Redundancy Manager
Pronounced “m-eh-rs-ee-t”, with the first “e” being like the one in “explain”, mersit is a simple Python script (Python 2.7, because I wasn’t sure what libraries were available for 3.x nor if my servers would run it well) hacked together with some sections that definitely look like spaghetti code. The good news is, it works fine, and has been well tested, so if you study it in the “black box” way, there are no big problems with it.
The purpose of the script is to manage the DNS records of the website served by the group of synced servers, in this case, tny.im. It runs on each server, in a peer-to-peer fashion. The peers select a single master, that will monitor all the peers and manage the DNS as they go up and down, “deciding who’s on stage”, and all peers will check whether the master is up, and select a new one that will edit the DNS to “hide the master from the public” when it goes down.
I definitely want to open-source mersit at some point, but not now because it’s not ready for prime-time (see “spaghetti code”, above), and I want to change some things that will make it more general-purpose. mersit has been managing the live records for tny.im for the past week (it’s been peaceful).
Continuing our journey through the world of meaningful graphs, here’s another.
I have gone so far as to write a read-me for mersit (mainly for me to read, as I know I’ll forget how it works within six months). I think it’s best if I put the start of the read-me here, instead of trying to explain it all, once again:
mersit - Tiny Redundant Server Manager
Copyright 2014 tny. internet media
This version is customized for tny.im
This is a Python 2 script that manages a group of computers/servers/thin clients/machines in a network (local- or wide-area), by automatically executing actions when something relevant happens to one of the machines.
We'll call the "machines" "peers". mersit assumes all peers and the network are trusted.
The script is meant to be run directly on the peers that are to be controlled, in a setup where there is not a single point of failure. It is not of much use when run in a single peer; in the context of this script, a "group" only starts to make sense when it has over one element.
We'll refer to this script as "controller software" or simply "controller", and to the other software that runs on a peer and which is to be monitored as "application". The controller is made to run unattended, even though it accepts commands (issued by an "operator") to trigger certain behavior manually.
The "something relevant" mentioned in the first paragraph consists on one of these "events of interest":
- A peer goes "online", that is, it is reachable by other peers and reports the status of its controller software as "OK" or "ready";
- A peer goes "offline", that is, it is either not reachable by at least some peers, or the controller is reporting its state as "not good" or "not ready";
- A peer becomes good-for-work (GFW), which means, that the application is functioning properly and performing its function (such as listening for incoming connections, data to process, etc.);
- A peer becomes not-good-for-work (NFW), in which case the application is not functioning properly, is too busy to perform its function (over capacity), or is otherwise unavailable.
Each peer works in a given "domain", which is the group the peer belongs to. The domain is specified by a name and secret which act basically like a username and password pair. Peers will only communicate with other peers of the same domain, that is, peers where the domain name and secret are the ones the controller is configured to use. The domain acts as the authentication element; an external party can not join, communicate or perform actions in a domain unless it knows the name and password used by the peers of the domain.
(Please note, that communication between peers is not encrypted by the controller - it goes completely plain-text over the network. It is possible to secure the communication between peers using external tools; such secure functionality goes beyond the scope of this software. The "domain" is simply a basic authentication system, implemented using HTTP authentication, to ensure that peers of a certain group don't start talking with peers from other groups. The basic authentication system is enough to protect against the casual script-kiddie, but by no means adequate for protection from a malicious party in an untrusted/open network)
The controller on each peer must know _a priori_ (i.e. before it starts) about where to find at least some of the controllers on other peers. Peer discovery doesn't happen automatically, however, once a peer's controller can communicate with another controller, it will add every controller in the "contact list" of the latter to its "contact list".
Imagine the following situation: you have peers A, B and C (and their controller software). The controller in A only knows about peer B. The controller in B only knows about peer C. If you start the controller on peer A, then start the controller on peer B, peer A will tell peer B about its existence, and peer B will tell peer A about the existence of a peer C (independently of peer C being running/reachable). However, if the controller in A knew about no peer (other than itself), it would never find peer B or C even if their domain settings all matched. Even though a big domain can be bootstrapped from just two peers, to ensure good operation, all controllers should know about all peers. This way, if the controller on a peer resets for some reason, it will have a greater chance of reaching another peer.
The "contact list" is the list of "known" peers. The controller keeps three lists of peers in memory: the "known" peers, the "reachable" peers, and the "GFW" (good-for-work) peers. The list of known peers is initialized from the source code's configuration section when the controller starts. It then proceeds to see which peers are "reachable", that is, can be reached through the network, are in the same domain (not being in the same domain gives the same effect as not being reachable over the network) and have their controller software report its state as "OK".
This initial status checking includes the exchange of some other information about the controller. Once this initial peer identification is done, the controller enters a monitoring loop where it will keep the contents of the three lists up-to-date. The controller keeps running this infinite loop throughout most of its lifetime. How the lists are kept up-to-date and what happens when their contents change is something that depends on the current controller mode.
There are two possible modes for controller operation: master and non-master. There is exactly one controller in master mode per domain, and this controller is usually called "the master" (the master peer has the controller in master mode). The differences between the modes are mostly related to what happens in the monitoring loop, but before going into those differences, it is important to understand how the controllers decide which peer is the master peer.
When a controller starts and there are no reachable peers, it promotes itself to master, since there must be exactly one master per domain. Later, when another controller joins the domain (either because it started or because it went online after e.g. a period without connections or power), it checks which peers are reachable from its "known" list and "asks" them which is the master peer. Every peer should reply with the same peer, in which case the new controller assumes that peer is the master, and informs the master about its existence, to account for the fact that the new peer may not be in the master's "known" list.
However, and especially on domains where not all peers initially know about every other peer, it's possible that a "head split" occurs and there are two masters in the same domain. Imagine a domain where there are four peers D, E, F and G. D only knows about E, which in turn only knows about D. F doesn't know about any peer, and we'll leave G aside for now. All peers are offline.
The D controller starts up, sees it can't reach the only peer it knows (E), so calls itself master. The E controller starts up and reaches D, D says it is the master, E assumes D is master, all is fine.
The F controller starts up, sees it can't reach any peer because its "known" list is empty, so calls itself master and sits quietly waiting for someone to contact it, which in turn would let it know about more peers.
We now have the following situation ([M] represents a controller in master mode, --- represents the knowledge peers have of each other):
| D[M]-----E F[M] |
Things could be like this forever, and no conflicts would occur - however, this is probably not a domain you want to have, since F doesn't know about any "event of interest" related to D or E, and these two don't know about any events related to F. In this situation, D--E and F act like separate domains.
Assume that G is a peer which knows about D, E and F, and that its controller starts up, contacting D, E and F. The first two will agree that D is the current master, but F will disagree and say it is the master. At this point we have a conflict. There are many ways to solve this, including some form of "voting" (e.g. the peer the largest amount of the peers say is the master effectively becomes it), but mersit solves this in a simpler way.
The controller checks that everyone in the domain agrees on what peer is the master on every iteration of the monitoring loop. It does this by "asking" each peer in the list of known peers who is the master. The first peer asked is free to reply with any peer. The ones that are asked next must agree with the first one. If not, the controller that was doing the loop tells each disagreeing peer that the actual master, is the one from the first peer's reply. It is possible that a minority is asked first, and thus everyone is forced to "change its opinion" to that of the minority. This is not a problem - mersit assumes all peers are trusted. Note that it can sometimes take some iterations of the monitoring loop for all peers to settle on a single master, because two (or more) peers may be trying to "change the opinion" of the other peers to different masters. This is not a problem either, because even if this kind of concurrency conflict happens once or twice in a row, it will stop happening as soon as one peer is faster than the other to tell everyone (including the other peer(s) that are trying to "change opinions"). What matters is that in the end, every peer knows about all others, and there is a single master. In this case, it could be D:
| D[M]-----E-----F-----G |
If the master becomes unreachable, or its controller stops working, the other peers will also find themselves a new master, by sorting the list of reachable peers alphabetically and choosing the first peer in the sorted list. Of course, if for some reason the list is not consistent across peers, the peers will try to "convince" others to settle on who they "think" is the master as previously explained, until everyone is set to the same master.
Being the master essentially changes what happens in the monitoring loop. When a controller is in master mode, it is responsible for updating the list of "reachable" and "GFW" peers, by checking which peers are reachable (both in terms of network and in terms of functioning controller) and which have the application in a working condition. If there are changes in the lists that indicate an event of interest, it runs the appropriate handler. If, for example, a peer becomes NFW due to a problem in the application, it will stop being in the GFW list, and the handler function for when a peer leaves that list will be run with the peer in question as the argument. If the master becomes unreachable (network error, controller error, etc.), a new master will be found, as explained in the previous paragraph, and the new master is responsible for running the handler with the previous master as argument.
When a peer is not master, it won't run any handlers for events of interest, and it is not responsible for updating the "reachable" and "GFW" lists - it will retrieve these from the master. The controllers on all peers need to keep their lists up-to-date, sharing a "vision of the domain" similar to that of the master, so that any peer can become a master instantly in case of necessity, without having to spend time performing checks on all peers and ensuring it has the best-and-latest list of "known" peers.
The operator can manually tell a controller to become the domain's master. When the appropriate command is issued, the controller will send a command to every other controller instructing them to switch to the new master. This command may not always have an effect in some controllers, because while the first controller is sending the commands, other controllers are seeing if everyone agrees on who's the master, and issuing the same commands with another master in mind. This is a sequence of events picturing the situation, in a domain where there are three peers H, I and J, and H is the initial master:
1. Peer H checks that every controller agrees it is the master (all agree);
2. Peer I checks that every controller agrees H is the master (all agree);
3. Peer J checks that every controller agrees H is the master (all agree);
4. Operator issues command for peer I to become master;
5. Controller on I assumes it is master, starts sending commands to other peers;
6. Peer H checks that every controller agrees it is the master, before the message from I that I is the new master can get to H;
7. Peer H finds out I (and possibly others) don't agree, sends them commands to change the master back to H;
8. Peer I changes master back to H;
9. Peer I checks that every controller agrees H is the master (all agree);
10. Peer J checks that every controller agrees H is the master (all agree);
If the master doesn't change when the manual command is issued, it's a matter of trying again. Most often, this kind of concurrency problem does not occur, and even when it does, it does no damage. While it is true that mersit could detect this situation and keep issuing commands automatically until the decision takes effect, we chose to not make it this way to allow the human operator finer control.
The primary focus of mersit is to monitor a distributed application. The master checks if the application, or part of the application, running on a certain peer is in working condition by asking that peer's controller about the state of the application it is monitoring. In turn, this controller runs a function, defined by the mersit user in the mersit source code, that should check the application and return True (if application OK) or False (if not). This can involve, for example, making a HTTP request to a HTTP server in that peer to verify it is working. The controller then communicates the status of the application to the master (which may be itself). All this shouldn't take too long, especially when the domain has many servers, as only one peer is asked at a time. If checking the status of the application typically takes over one second, it is best to store the last known status in a variable, and update that state periodically in an asynchronous manner that may be external to the mersit script.
The part related to DNS records is not explained on the read-me, because it is related to the handlers (which each mersit user would customize to the specific needs of the system – as I said, I tried to make it a general-purpose script). Sounds interesting? Feel free to ask questions, or point out problems, in the comments.
tny.im and this website were down for about 42 hours, starting on June 29 at 03:16 UTC.
The problem? BlueVM’s S19-NY server went down, taking with it the server I have/had there (and which I paid for a full year!). Other than this outage, the service had worked fine for three months, – fast network, full resources availability – since I bought it.
S19-NY is still down, without any ETA for it to come back. There’s no information in what conditions it will come back, or *if* it will come back (with the previous contents, at least). BlueVM staff is pretty much unresponsive, other than a guy who sometimes hangs on IRC and says he can’t fix the KVM instance because he doesn’t have access to it.
Of course, I no longer recommend BlueVM and I don’t plan on renewing the server I have with them.
The “solution” to put an end to two days of downtime, was to buy the cheapest SecureDragon OpenVZ server, (OpenVZ! so hard to live without my beloved KVM, I can’t even use davfs2 because there’s no fuse module!) and restore the backups I had (from four hours before the BlueVM instance went down). This has been done, except for the HTTPS certificate of tny.im… that alone is another story:
As I tried to retrieve the existing cert from StartSSL (because, stupid me, automated backups were not copying everything SSL-related, and I didn’t save it locally), I found my authentication certificate had expired. This basically means my StartSSL account is lost, unless I create a new account and ask their staff to link it to the old one. They probably won’t do that without a payment and some ID checks so… out of question. I guess, if the BlueVM server doesn’t come back, that I’ll just create a new StartSSL account and generate a new cert for tny.im. There’s no security issue with this, as the previous certificate has not been compromised (unless BlueVM is collecting certificate private keys from inside their clients’ machines…) and so it doesn’t need to be revoked.
To conclude the HTTPS point: tny.im has the HTTPS service unavailable for now, until I can retrieve the existing certificate from the previous server, or until I get a new one.
Is all the fault in BlueVM’s side? Of course not… I could have lost my love for the money earlier, bought the SecureDragon VPS yesterday already and reduce the downtime by 24 hours. But since I had hope the problems on BlueVM support were just Sunday-related, I thought that by Monday they would have it fixed. They didn’t.
On related news, I’ll take this downtime and new server acquisition as the motivation for setting up a new advanced and redundant system, so that if one server goes down, tny.im (and possibly this blog too) will continue to operate as normal. I have two servers already (assuming the BlueVM one comes up), and I plan on developing a system where firing up new instances of tny.im on any empty server will be really easy. The system will be always prepared to lose any server at any point, without data loss, and restore full service within five minutes. That way, I can add less reliable hosts, perhaps even VPS trials, to the redundant system. This also allows for scaling the service as needed. Sounds ambitious? Of course it won’t happen in a week, but I have the full summer to develop and test the system…
Why don’t I just go with some SaaS that supports scaling? Two reasons: the price is too high, and the tny.im software is not coded in a way that’s compatible with these services. Let me remember you, that while not exactly being a CGI script, tny.im is not coded in one of those fancy modern languages, and even though PHP is not exactly outdated or unmaintained, the quality of the code can make it pretty horrible or pretty good. And the code is… not perfect – it doesn’t use any popular framework, it is based on YOURLS and has many, many
hacks feature additions, plus a… very close relationship with the database.
Let me finish by saying that downtime of this kind is something to be expected if I were still hosted by FreeVPS and the like. But believe it or not, on FreeVPS and other sponsorships I’ve never seen a customer service as bad as the one of BlueVM (and it’s hard to remember an outage as big as this one). It is definitely not adequate for a paid service. In addition to S19-NY, they have many other instances down, with similar or worse downtime. The admins don’t appear to be online or reachable in any way, even by other staff members. The latest news/excuse on the S19-NY situation is that IPMI is broken and they are waiting for the provider to fix it… now tell me, does this look like a serious company, or some poor-man’s sponsorship?
EDIT: The BlueVM server is still down. tny.im is now hosted by three servers with round-robin balancing. HTTPS service was restored with a new certificate.
EDIT 7/7/2014: I forgot to update this post in time, but the BlueVM server has been up since three days ago. But I only got to know that the service was restored thanks to a friend of mine, because they didn’t reply to my ticket to inform me about it. Anyway, I don’t plan to renew this server, and BlueVM lost me as a customer (except for some really cheap deal which I’ll use as a personal/development box, and never in production).
On related news, Mirasm – the Tiny Server Redundancy Manager – is mostly finished, only needs some more testing to be put on production servers, managing the new tny.im redundancy system.
Looks like >90% of the Internet users spend >90% of their WWW-using time on the same 0.001% of websites.
It’s no wonder why plans for a truly decentralized Web never seem to get much traction… how could them, in a time when users’ habits are truly centralized? In fact, I believe the main problem is on the way users use the Web, not the technology iself.
As for the Internet, it is already (comparatively) decentralized.
Do users always know what’s best for them? The past has proven that is not the case…
Just in case you didn’t know/notice, this website and tny.im are no longer hosted on a VPS provided by FreeVPS. Due to issues on the node where that VPS was hosted, that made it become network-unreachable, I needed to buy a new server to avoid a longer downtime.
I ended up buying a KVM-powered VPS from BlueVM because they accept Bitcoin and had a good special deal at the time. It’s my third day with the server and so far, so good. It has better specs than the server I had, and after config optimization tny.im is now able to handle more concurrent requests than it did before. I will keep readers updated and post a detailed review of the server, later.
Moving away from sponsorships is be better in the long term (less worries, proper support…), even if it means that tny.im is now unprofitable (temporarily, hopefully).
If you watch other websites, pages and forum threads of mine, you may already know about this, but just to make sure you don’t miss it, v1.3 of Utilities is out. Download or more info.
Regarding the problem of Casio Prizm calculators that hang on the power-off screen and then die when rebooted, I have the feeling that we [the Casio programming enthusiast community] have discussed this subject way more than Casio ever did – in the end, it may be easier and cheaper to fix or replace, say, 1% of the calculators they sell, than to pay lots of working hours to discuss and fix hard-to-reproduce bugs, delaying the production of newer models (in the end, nowadays people buy a calculator because it’s fancy, not because it will survive the apocalypse ).
To all the users with dead calculators… I know your situation is bad (especially if you didn’t have backups of the data in it), but just send the f***ing thing for repairs before the three year warranty expires. Make sure to explain how it broke, maybe they will fix it some day. And when the warranty expires, if you still need a graphic calculator, well, buy a new one (eventually not from Casio?).
In case you haven’t noticed yet, the version 1.2 of the Utilities add-in for the Casio fx-CG 10 and 20 (known as Prizm) has been released today. More information and download on this page.
Following my announcement on Twitter about the new release, it got retweeted by the official Casio Prizm Twitter account. This is a move without precedents on their part – no 3rd party add-in had ever received the slightest official public recognition. Most likely the social media marketeer retweeted my nice tweet without really knowing what he/she was doing.
However, this had little impact. The tny.im shortlinks you see received less hits by the time they retweeted, than by the time I originally tweeted – despite @CasioPrizm having over the quintuple of followers I have.