Rant: Linux vs Windows (and a little Open Source vs Closed Source)
Collegue: “… we can argue about advantages of Linux vs Windows for a long time, but there are constrains that require use of Microsoft products due to support of various software and hardware. A very simple example of ASP dictation usb device… that only has windows client. There are limitations, design constraints and manageability concerns. Neither you have an much experience of managing Linux based network of hundred nodes.”
I’ve managed a handful of nodes for both Linux and Windows. Currently I’m fighting to keep control of this Windows 100 node beast, but the different nodes have different issues, they all have their own installations of operating system, their own installation of office and PDF and antivirus (“client server antivirus”, but it doesn’t behave as well as advertised), and each one might have a different version of each of these necessary parts. A thin client system has none of the hassle with this. Manage one image with one set of tools and utilities for the majority of workstations. The handful that need extra care are then in a smaller classroom and can have more 1 on 1 time with teacher.
- Linux had “Data Deduplication” in 1993 (ext2 file system with hard links), Microsoft in 2012.
- Linux had “Distributed File System” using nfs(1984), Microsoft in 1996.
- Linux had a 64bit version in 1995, while Windows was just moving from 16 to 32 bit architecture.
- Linux was the first operating system to run on the Intel IA64 processors.
- Linux or another Unix are the primary operating systems for all supercomputers.
- Linux is the operating system of choice for render-farms for all those 3d movie special effects we love.
- Linux is the underlying foundation of all Android devices.
- HP Sells servers with Linux, certifies and supports it.
- Google uses Linux primarily for all of its servers and workstations.
- Texas Instruments uses Linux workstations primarily.
- MacOSX is a Unix-based desktop operating system, so the many dozens of businesses in San Francisco that are Mac-only are Unix-only. The MacOSX desktop (window manager, application launcher panel, etc) have been open sourced and are available for Linux.
A move to Windows7Professional or Windows8 will cost at least $10,000 before including a newer Office, unless we cheat the licenses. This multiple thousands of dollars has to come from somewhere. I don’t want to cheat, and I don’t think Microsoft deserves that much money for all the hassle they put their users through just to keep using the computer. Windows is behind Linux in stability, behind Linux in security, behind Linux in maturity, behind Linux in features, seriously starting to fall behind Linux in usability, and only ahead of Linux in a few hardware points due to Microsoft throwing their monopolistic weight around with threats, litigation, and everything BUT competition on merits.
It would be fascinating to see what would happen if a law was passed that simply said that computer vendors could no longer bundle an operating system pre-installed on the hardware, and bundling OEM pricing scams schemes are no longer allowed. Then every new computer would either come with a $200+ add-on for Windows or a $0 add-on for Linux (As opposed to now, when the OEM pricing allows Dell to bundle Windows for ~$35 a machine, while retail price is $200, and that OEM price is just hidden in the price of the machine) Imagine what would happen to this monopoly if they actually had to compete to get into people’s homes!
“Let’s see, I can buy Windows for $200, then Office for $200, an antivirus for $60, a PDF writing utility for $300, Photoshop for $700… that’s $1460, or I could go open source and get Linux for $0, that has LibreOffice for $0, doesn’t need antivirus, LibreOffice includes PDF output, and theGimp for photo editing for $0, resulting in a total of $0 added to the cost of the computer. I could double the machine’s CPU, RAM, and hard drive space, use higher quality components, and still not be more expensive than the closed source software path.”
Switching to Open Source systems is important to me, and I believe in the end it will be a benefit to any company in terms of stability, expandability, supportability, and both up-front and long-term cost savings.
Email for Business
The more I work with how different people use their email, the more I’m coming to realize that, for the vast majority of businesses and employees that use an internal email system, typically based on Outlook, the Email client is how people are keeping their files organized.
From an IT standpoint, this makes no sense. The email client wasn’t made to file things, the protocols weren’t even designed to hold all those attachments in the first place! Programatically, it makes little sense, as every email client has it’s own methods and limitations on the email storage system, while the normal file system does not. From a Desktop Technician standpoint, it makes little sense, as those email storage systems could get corrupt and THEN where are you? Plus backing those things up is a major undertaking, especially when massive chunks of the data are no longer really needed, they are just taking up space and collecting dust.
But from a USER’s standpoint, it is exactly what they need. It is simple, organized, keeps track of dates, times, replies, versions, keeps things in folders, automatically sorts incoming information, displays a preview as you browse through, and you can keep it all in one place.
The problem is that while the email client designs today ALLOW for this pattern of use, they aren’t DESIGNED for it. They are designed for email, to organize email messages, expected to be mostly text with occasional attachments. The design is not fitting the use, and not even the big new changes in major retail offerings for hundreds of dollars are moving in this direction. Most developers look at the prospect of an email client and think, well, obviously: email. Users and Management look at the prospect of email and think: documentation, tracking, records of conversation, simple filing of important information.
This dynamic departure between tool and use is not uncommon for humans. What surprises me is that the tool hasn’t been altered toward this use pattern. Not foundationally, anyway.
So my mind races to start designing a new email client, from the ground up, with several key points that are not-designed-for-but-used in mind. Directory structure, not some strange mail-store file format, compressed (always), possibly encrypted. “Local” storage mechanism compatible with network shares or nfs mounts. Standards based so that the files, attachments, etc., can be accessed without the email client. Additional files added to the store to keep things on track and smooth, like an XML index relating message files to their attachments. Pseudo-data-de-duplication by comparing new attachments to existing attachments in the folder, and keeping only one of a group of duplicates. This might sound like one of those “that never happens”, but then you may not have seen extended conversations where the same attachment gets passed back and forth a dozen times, but was only valid the first message, users just don’t know not to keep the attachment in the next reply. Threading data for conversations can also be kept in this XML index.
The concept starts getting exiting for me when you look at it and say: you don’t need to have an email message to manage files in this system. Documents or other files, unrelated to emails, stored in folders with XML metadata for times and other useful searchables, inside the “mail store”, because that is where the information “belongs“. A project has it’s own folder, files specific to that project should be there, regardless of messages in or out. Adding a nice transparent versioning system would round out the functionality that is already being forced into the mold of an email client.
An argument might be that it could get confusing with files and emails and whatever-else, but my counterpoint to that is: this is already a pattern of use. The tool is behind the users.
But my “Grand Plan Alert” is ringing in my head, telling me that this idea is a MAJOR undertaking, even if I start with an existent open source email client.
Features:
- Spam Filters via External Blacklists and Internal Whitelists
- PGP Signing and Encryption (partly optional)
- Conversation Threading
- Extended Rule Sets for Filtering and Organization
- Automatic Attachment De-Duplication
- File Storage and Versioning in Email Folders
- Mail/File Store Compression
- Mail/File Store Encryption (optional)
- Calendars Linked to Email Folders
- Shared Calendars
- Contacts Linked to Email Folders
- Shared Contacts
- Email Lists Linked to Email Folders
- Shared Lists
- Transparent Encrypted Email Based Distributed Shared Files and Versioning System
- Server-based File Storage with IMAP
Yeah, then all I’d need to do is convince Microsoft to buy it from me AND give it away for free so everyone just has it. Would solve all my email headaches.
Enterprise Linux Musings
Frugality.
Frugality is what I’m faced with as both an obstacle and a philosophy in my daily work environment. Frugality and Microsoft don’t mix (unless you are cheating, but cheaters never prosper). I have to move 100 workstations off of Windows XP this year (I’ll skip the 8 servers that have to be off of Server 2003 as well this article).
100 copies of Windows 7 or Windows 8 Professional Upgrade: ~$10,000 (too small to be “bulk”, too large for this to make sense)
[Edit: Windows 8 Pro Upgrade: $200. So that's $20,000. ref: http://windows.microsoft.com/en-us/windows/buy?ocid=GA8_O_WOL_Hero_ShopHP_FPP_Null ]
Just from there, I’ve only got 9-12 of those systems that MUST remain Windows systems. Some of the “new” workstations we regularly get to replace the dinosaurs that regularly die off already have Windows7Pro licenses, so I can likely handle that number. I believe I’ve got 2 unused Windows7Pro licenses right now, might get more this week, but I digress.
Turning all of these older, low-power workstations into thin clients is looking more and more like the right move. All but the dinosaurs already support PXE, so the short story is: move DHCP to a Linux Server that hosts a PXE boot ROM, and pull the hard drives from the workstations. A little user training later, and the WindowsXP problem is solved. Perhaps a ram upgrade or 12 to make things run smoother. Maybe instead of yanking the drives just quick-wipe them and make them swap spaces, since I have to touch each workstation anyway. I might even be able to make a DSL USB boot stick that does this prep for me, wouldn’t even have to type much.
But the back end is a little more complicated. I haven’t quite mastered getting a Linux server to authenticate via LDAP to Active Directory. I haven’t built a test Linux desktop environment for the users to have once they are connected (because I haven’t got past the login part), and Linux RDP is still not compatible with our current EMR system (due to our EMR vendor’s addition of a proprietary File Browsing and Scanning channel to RDP).
I’d like to have a sane server environment that self-configures some things like Network Shares based on Active Directory information, using the logged-in user’s credentials. I want to have several standard programs that mirror the function they currently expect from a workstation: Email with Shared Contacts, Personal and Shared Calendars, Web browsing, PDF open/read/print/fax, Document open/read/modify/write/print/fax (“Word” and “Excel” files mainly), easy shortcuts to the File Sharing system with correct permissions (from Active Directory), and the RDP connection to the EMR system.
The Thin Client method works better than the Linux Desktop method in this case because the users are connecting to a central server that can force-reset settings to keep things in control, such as keeping unwanted plugins out of Firefox and Thunderbird and keeping wanted plugins and settings set properly. I haven’t decided solidly on Thunderbird, as Evolution looks a little closer to what people used to Outlook would expect, but I don’t know if Gnome is light enough for an always-remote environment. I do need to keep the network traffic down where I can. KDE is a Linux favorite and often said to be the official standard, but again I haven’t really dug into Desktop just yet.
So there is a lot of development that needs to happen, and if I’m doing it I need to do it along with all the other things my job demands, which will make it slow. The best possible outcome is a system wherein a server is installed, the workstations are remotely switched, and everyone comes to work the next day to a new system whose interface is not so unfamiliar as to interfere with workflow, with the benefit of being stable and fast as well.
Linux, Modems, and Faxes
Ever since they came out, I’ve hated “WinModems”. Even back in the first days of them, with Windows 95, the concept frustrated and annoyed me. The grand idea of these ‘modems’ was that the computer’s processor is enough of a powerhorse that the manufacturer doesn’t have to put expensive digital signal processing hardware into their ‘modem’, they can just off-load that task to the computer’s main processor. Of course, that requires ram and takes up a significant portion of the CPU’s abilities (especially back before gigahertz speeds), and obviously it takes a pretty beefy software driver to turn the CPU into a DSP to drag the digital signal forth from the squeeky analog noise on the phone line.
That’s where the “Win” part comes in. If you have a “WinModem”, it is unlikely to work in anything other than Windows. In fact, when Windows 98 came out, I had many customers show up because their WinModem driver for Windows 95 didn’t work in Windows 98. The fix was obviously a downloadable new driver. Yes, downloadable, but the modem doesn’t work… right. Sure, technically minded people like myself looked this kind of thing up BEFORE upgrading an operating system, but most people just pop in the upgrade disk and click OK. There was a trickle, nowhere near the same number, when Windows 98se came on the scene. I always recommended people find and spend the extra few dollars to get a hardware modem. The real thing, no drivers needed, shows up as a serial port and just works. My go-to company was 3com/USRobotics.
This past week I’ve been spanked hard by what should have been a little problem, except that instead of Windows Server 2003 throwing ERRORs about it’s inability to send ANY of the outgoing faxes, it just gave warnings. Warnings don’t ping the monitoring scripts, so I didn’t solidly know there was a problem until Thursday. Nothing had been sent since Monday. Microsoft’s Fax Monitor and Fax Administrator panels (in Server 2003) don’t do any counting for you, other than how much is in the queue (no “X jobs selected” when you select a chunk of them), so I don’t have a solid number, but for the week I’m sure it is between 500 and 600 faxes. It’s been sending all weekend, it hasn’t caught up yet.
I can’t help but feel that there has to be a better faxing alternative. Yes, I understand that “most people” don’t use faxes anymore, just scan and email and be done with it. But email is not HIPAA compliant, has never been considered secure or private, and so sending confidential information via email is frowned upon at best and illegal at worst. Fax machine to fax machine is considered as secure as the physical machines on both sides of the connection. Our EMR software can send a fax directly, and this is what our clients expect. So “just don’t use faxes” is off the table as a solution.
I thought, why not a Linux fax server, so I don’t have to deal with the constant windows security updates, antivirus updates, and these awful USB glitches. I’ve never had a USB device just disappear from a Linux box, except once when a wireless mouse USB dongle died… but it was dead, wouldn’t work anywhere after that. These USB modems in this server just… glitched away. They are working fine now. Typical Windows “reboot and deal with it” kind of problem.
But the modems are WinModems. And just about every reasonably priced modem on the market now is Windows Only. The handful I’ve seen that list Linux don’t have such good reviews, some to the point of “It says it supports Linux, it doesn’t, don’t bother”. So find a hardware modem… back in the day they were about $30 when the WinModems were $15. Today the WinModems are $10 and the hardware modems are $80. *boggle*
I could very nearly build one from chips and a circuit board for half that. Especially since I only need the 14.4k maximum Fax speed.
I wonder if a Parallax Propeller could do it with software alone for $8 + crystal, eeprom, board, FT232RL chip, RJ11 jack, case and usb cable. The thing has more than enough speed and high enough analog sample rates, just a question of firmware. With all that stuff, I’m probably looking at around $20 at one-off volumes, minus firmware development costs of course.
An open source, open hardware modem for everyone. Would be nice.
Goals and Obstacles
Ever since installing Linux for the first time in 1999, and switching to Linux as my primary home desktop operating system around 2001, I’ve pondered and *boggled* at the vast amounts of money still being spent on bloated, insecure, and unstable Microsoft products. This year, I am faced with near 100 workstations running on Windows XP, that need to not be running on Windows XP by this time next year. I could do the typical “IT Guy” thing and say that what needs to happen is a bulk purchase of 100 licenses for around $10,000 , and then upgrade 20-30 of the workstations to be able to handle the new operating system for another $5,000 , and that’s the end of it.
But why should it cost so much? $15,000 for nothing more than continued function, tiny increases in hardware support, arguably a decrease in usability and user job performance due to drastic user interface changes, increased IT support effort due to running Microsoft’s latest “release-beta” and user expectations no longer matching the Operating System, on top of all of the server-side changes already in the pipe. Specifically, in order to support Windows7 or Windows8 network printing, I must upgrade the file/print servers to Server2008 or later, because Server2003 won’t distribute printer drivers for 64bit operating systems other than XP64. And I DON’T want to deal with trying to support a mixture of workstations with 32bit here and 64bit there.
All this, ALL this, for the primary purpose of… connecting to the terminal server via RDP, with secondary web, email, and document processing, including PDF creation, printing, and faxing.
…um, I’ve seen Live CD Linux distributions fit that job description, on hardware much more ancient than our oldest workstation.
For free. (Some would argue time and effort, but I’d have to spend that either way. With Linux, I just don’t have to spend the MONEY.)
My brain screams, “I CAN SAVE $15,000 THIS YEAR!”
And that doesn’t even touch on using IMAP/Thunderbird instead of Outlook/Exchange, which would save ANOTHER $10,000.
$25,000 off the bottom line. Open source tools with massive community support (also free). Commercial support available for a price, but not likely to end up costing that much. And there isn’t a major monopolistic power pushing to make us spend all that AGAIN in 3-5 years.
…
But “everyone” is running Microsoft Windows programs, and Microsoft Office programs. Our primary application adds a driver “channel” to RDP for file browsing on the local workstation across the RDP link, that is provided by the application vendor and they don’t support Linux. Our alternate vendor’s product isn’t ready for our needs, and probably won’t be for six months to a year, realistically. Many programs are in use that are required, that don’t have viable Linux alternatives, such as Quickbooks, Millennium (our payroll company’s report program), and a handful of other programs that are specific to a handful of people and their duties.
Standard Users, as I’ve come to call the people who’s job is roughly 90% RDP and 10% email/PDF/fax, could fairly easily be migrated to Linux desktops or even Thin Clients. This represents the majority of the users in the company. But, in order to keep the other users, which are higher-level users, typically management, running the way they expect, they still need Windows and their special tools. Even if I managed full transition, I would need at least Virtual Machines running Windows for these tools, with the accompanying Windows support structure, updates, antivirus, and backups failing due to fighting with volume shadow copy (though shutting off a Virtual Machine and backing up the .VHD could be easier than on-VM installed backup software, though not quite as flexible or effective for those user-side “oops I deleted something” restores).
The vision: ZFS based SAN plus a nice Linux Thin Client Terminal Server that can PXE boot a workstation, ZFS replication between clinics for common documents, Unix Group based file access rules, LDAP Authentication for central User and Groups management (possibly retaining Active Directory). Server based printing, NAS for home directories, Thunderbird+IMAP+LDAP for email and contacts, OpenOffice or LibreOffice for documents and PDF viewing/annotating/creating. Faxing… that one is a little harder. Putting Fax into the obstacle column with Quickbooks, Millenium, Microsoft Office documents with Passwords, Adobe PDF files with Passwords.
But my cohort seems so calm and reasoned about sticking with Microsoft and spending an entire year’s salary (for some people) on software licenses that effectively amount to: continued use. He says that people wouldn’t pay for it if they didn’t have to, or didn’t see the value in it. He says that there are many reasons that Microsoft is the major player, they know how to support business, they research and plan and make products and decisions that help businesses succeed. He says if they didn’t, people would vote with their money and Microsoft would be out of the picture.
Nevermind that most end users don’t even know that Linux exists, that a free alternative exists. Nevermind that MAJOR enterprise businesses like IBM, Oracle, Texas Instruments, Google, Symantec, and Amazon rely HEAVILY if not ENTIRELY on a Unix/Linux infrastructure. Nevermind that Open Source is proven to be resistant to viruses and malware by foundational design. Nevermind that Microsoft has been proven to both legally and illegally attempt to control and eliminate Linux as a threat, by Fear, Uncertainty, and Doubt marketing campaigns, funded “research” publications, and blackmail under the guise of patent litigation. If I could, I’d boycott Microsoft flat out, but I can’t, so I’m doing the next best thing: slowly removing them from power over my life and the lives of those around me.
What I want is an Enterprise Level Linux Suite, with centralized workstation and user management (LDAP), email management, LDAP shared contacts, user and group level Internet whitelist/blacklist, distributed document library, replicated user profiles, Site-to-Site VPN, Remote VPN, seamless Printing/Faxing, seamless backup/restore, capable of supporting Thin Clients, Light Workstations, Heavy Workstations (including Windows workstations that “think” this is a “Windows Domain”), Physical Servers and Virtual Servers, with Virtual Server high-availability and redundancy options, and Physical Server local and remote clustering options.
None of this is outside of the current state-of-the-art, but there just isn’t a single-package Suite for it.
Yet.
* Any trademarks presented are property of their respective owners. Any opinions stated here are those of the author and no other entity. Items stated as if fact without reference should be regarded as opinion.
Belief vs Career
I’ve been having this unsettling feeling for several weeks at work. I haven’t really been able to put a finger on it, and every logical argument I’ve come up with has not held. The product is Microsoft Sharepoint 2010 Foundation. The purpose is to replace the (admittedly quickly thrown together and not nearly as friendly) LAMP (Linux, Apache, MySQL, PHP) based solution I built a few months ago, and add functionality and ease of use. I worried that whatever limitations Microsoft has placed in Sharepoint 2010 Foundation will be run into quickly, and will cease to be free, but the consultant who is pushing the business this direction assures us that with the intended data and use patterns, it’ll be 3 or 4 years before we hit the storage limit, and there isn’t an artificial user limit in the software.
So it’s free.
And it has more features.
And is easy to use for the end users.
And it’s connected to Active Directory so a lot is automatic.
And it’s pretty easy to build new pages and tools for the end users.
So…
Why does it still bug me so much?
I’ve been chewing on that question for weeks now, and I think I’ve finally come up with the truth: I believe in open source technology and moving toward open source solutions, and switching to a Microsoft product is a step in the “wrong” direction. More specifically, a step in a direction that is against my beliefs. As an atheist, I am flabbergasted that I would end up running into a negative situation in IT that would boil down to my beliefs.
So, this chaffing at my “soul” is likely similar to many other religious belief violations that occur in other places of work in other businesses. The problem is that I don’t think I can get away with being a “conscientious objector” and still keep my job, unless I can come up with both A> a similar or better option from the open source community and B> a real technical reason not to stay with Sharepoint. Which means that I get to bite the bullet, learn the thousand layers of Microsoft (re)rebranded technobabble, with their hundreds of clicks through dozens of screens to get the things done that I need to do, find and buy the books and take the classes and pay for the certificates that prove I know what I’m doing with this tool (because unlike the open source community, money spent on books and classes are almost the only way to get this information), wasting time on and putting money into a business structure and model that I detest. All the while (because I’d go insane if I didn’t), researching, contributing to, and/or writing my own replacement of said tool.
… in my “spare” time.
I’m not sure I’ve got the energy for all that, but I’ve got to try. I have to focus on the eventual future where I can uninstall this thing, and keep it in mind as the sunny day at the end of the tunnel.
Political Correctness Offends Me
I’ve forgotten now how the line of thinking started, if it was a misstep in the morning radio show I listen to on the commute to work, or an overheard joke, or an oddly phrased commercial, but the thought, “Wouldn’t it be ironic if someone were honestly offended by political correctness?”
Yes, that would be sweet irony, someone being offended by something that is by it’s definition meant not to offend. But then I recalled all the pointed jokes, all the disdain, all the frustration and aggravation over the years centered on “PC”, and I realized, I AM offended by Political Correctness. Now it is layered irony.
The trouble is, even though there are many people who are obviously offended by the existence and expectation of PC terminology, myself one of them, I couldn’t readily say why. Why does this idea of putting forth some effort of vocabulary in professional and public speech and writing cause within me the revulsion it does? With that question, I did some digging in my head and found: wounded intent.
What I say, what I write, even what I sing (when I sing), I intend no offense. If someone takes offense at something I say, it is their effort, their intent to be offended. Whether they are looking for an excuse to react for attention, an excuse to argue their issues or opinions, an excuse to start a fight, or just an excuse to play victim, they are searching what is said and written for something to be offended by. How much effort should I expend to attempt not to offend someone who wants to be offended?
The Politically Correct answer is “as much as it takes”.
The fair answer is: nothing extra.
It isn’t my job to protect other people from their own issues. Intent is key, I don’t intend offense, so don’t take any.
Example: “Merry Christmas!” The intent is wishing a person happiness in the season, or on the day. December 25th is Christmas day, whether you believe in Christ or not, whether you have any traditions connected to the date or it’s name, or not. If you don’t believe in Christ/Christmas, and you AREN’T looking to be offended, this is as benign as “Happy Monday!” In point of fact, I often wish people Happy-day-of-the-week.
Example: “Happy Solstice!” Same basic argument. You can’t even argue that solstice is just a made-up day by some crazy religious group, since it is an astronomical event. Why not celebrate it? The days will start getting longer now! Happy day! Why take offense because I didn’t form it into something nonspecific and benign? The intent was to wish you happiness! Where’s the offense?
Instead of expecting everyone to tailor their vocabulary into some neutered form of pseudo-meaningful ritual semi-pleasantness, maybe we should expect everyone to look past the words to the intent.
Backing Up Windows Servers
One of the most important parts of IT is backups. Reliable, automatic, frequent, easy, and restorable backups. It SHOULD also be pretty darn basic, just set Windows Backup to run every night and it takes care of everything, right? Well, yes and no.
Windows Backup doesn’t do incremental backups, meaning that it’ll copy EVERYTHING, EVERY TIME. If’ you’ve already got a backup of 95% of everything, an incremental backup will only back up the missing 5%. Yes, I am aware that Windows Backup is SUPPOSED to do incremental, but it just doesn’t. It also doesn’t try very hard, it doesn’t retry if the Volume Shadow Copy step fails, it just gives up. So, the list of requirements of IT above, scratch off Reliable, Frequent, and Easy. Doesn’t look good.
I haven’t had first-hand requirement to restore from a Windows Backup in years, but the bad taste in my mouth of the loss back then still lingers. Three backup files, “Failed to restore”. No options, no rescan, no scan for recoverable files, just “failed”. Wow. BUT that was years ago and many updates and versions gone by, should be fine now, right? According to my research: no. Though it looks like the failures now are more often on the front-end, and it reports failure in creating the backup in the first place, but overall, the general feeling on the matter is that Microsoft does not have a reliable backup product.
So of course my brain leaps to the differencing, versioning, rsync-based backup I built for the company website. All I’d really need is to share the root drives with a backup user and I could rsync anything and everything, with resolutions down to hourly! BUT Linux rsync doesn’t retain the complex Windows permissions schemes. Well, that is both true and not. Linux has tools and modules that can work with Windows Access Control Lists some, but not completely. What I’d like is a way to read all of the ACLs on a file, store them with the backup, and re-apply them on restore. Another ‘sounds simple’ yet lofty goal.
The other side of the coin is to gather up the courage and tools to write my own Windows backup application. But my code-fu is still mostly missing, and I haven’t done Windows development since Windows 98 (around late 1999 I built my first Linux box, and danced with glee at all the FREE developer tools!)
Lastly, the ever dreaded “spend money” approach, but nobody agrees on any Enterprise-level tool for backing up mission critical servers and virtual servers. There are dozens of tools out there for it, and many with free trials, but none free, and “low” prices are mid-3-digits for the “licenses” I’d need for the three clinics I support. We have some Acronis Backup installed on some of the older servers, but the reviews of the newer Acronis Backup versions are not as good as the old ones, finding comments like “wish I hadn’t upgraded” and “digging up the old version to reinstall”. Back in the day I used Norton Ghost for workstation-level backups, but it doesn’t do a lot for hot backup of running servers and SQL.
Any suggestions?
Yay Mead!
The mead I wanted to release a year ago is finally out: Simple Pleasures hit the market this weekend! Winter Celebration, a mead with mulling spices fermented in, was also released this weekend and both went home with new friends this weekend. It is so gratifying to have so many people taste my take on mead and enjoy it enough to buy it. I’m sure it is a widespread “know that feel” among winemakers and beer brewers, the, “I spent a long time crafting this, for me it is a work of art, and now not only do I get to share it, but people are actually willing to part with their hard-earned money to keep more than the little taste I gave them!”
Such an awesome weekend.
Over at Lightheart Cellars you can see the meads available. Currently we can only ship within California. But it’s a start!
Yup. Been Busy.
It’s been a good busy, but very busy. Wine has been made, mead has been finished, labels have been designed and submitted for approval… the winery had to be expanded for the amount of wine made due to the amount of grapes harvested. We are still not going to use all the grapes available! Not for lack of desire, but for lack of man-hours and square-feet. Tons of positive feedback on the Winter Celebration mead, with many people eager to purchase bottles for holiday celebrations! Just need TTB’s approval on the labels so I can bottle it and sell it! So excited!
Meanwhile, the Ghia got my birthday present (1600cc not-destroyed engine) installed (only been 8 months), but it doesn’t run yet. I might just have to replace a good majority of the intake and electrical components with parts from the other engine, or new ones (like the distributor I fried, again. Did the exact same thing last time I removed/re-installed the engine, but to be fair it was 3 years ago).
At work, we’re building up to opening a new clinic and switching Electronic Medical Record systems, not quite at the same time. New servers, new firewall, new rack, new UPS, and I’ve got to make it all work with the old stuff. Some of the old servers still running are nearing 10 years old, which is like 100 in computer years. Last week I saw a 20gb harddrive still running on a workstation! Well, “running” might be the wrong word, maybe “still operational”?
Other than that, still at work, I managed to build a versioning backup system for the website, and I’ve been building a script to run on-top of the resulting backup data to make it easy to find and restore deleted files or find and rewind to previous versions of files. This is my first real dive into jQuery and my first step in probably 6 years into “Ajax” programming. Been fun brushing rust off of some places and welding in new concepts on the old foundation. I’m not sure how I feel about knowing that I really like border-radius. Anyway, once this backup thing gets to a usable point I may just release it to the world, helping others have the confidence of a good, automatic, easy-to-use, relatively efficient, backup system with easy-to-use recovery built in. The part I like most is that I can do the recover/rewind bit without the script, since the storage back-end is simple. I just wouldn’t want to spend the hours it would take to walk an end-user through the process.
And on the home front, the Daughter is still passing all her classes, and we are going to DisneyLand for Halloween. I’m going as the Mad Hatter, from the more recent one with Johnny Depp. It’ll be a blast!
When we get back I’m hoping the label will be approved and we can bottle the Winter Celebration mead and start selling it for the season.
Last, of course, this is fall, the clocks change not this weekend but next, closing out my vacation with a time change. Not sure if that timing is fortuitous or frustrating, but I’m leaning toward frustrating simply because Daylight Savings Time has always frustrated me. I’d rather drive to work in the dark and have some daylight in the afternoon when I get home. But no, I am not moving to Arizona: I really really like the San Francisco Bay Area. Digressing again, Fall: change in seasons triggers in me changes in interests and focus. Fall has historically led me toward flight and robotics as primary interests, but I’m seeing a trend of Project Flood as well in more recent years, because ideas come, and some are good, so I write them down and start planning them out… before they end up on a back-burner due to, well, life. One such project, ironically, was/is designed to help manage such “back burners“. lol
So there is a (not-so) brief update on the month. Now it’s almost time to get out of work and be on vacation for a week!