Many people and corporations are putting their data into servers hosted on the the Internet. Some, like Amazon and Google, have opted to use Linux servers. Others, like GoDaddy moved their parked domains from Linux to Windows, a dubious moved designed to bolster Microsoft's Internet server share at a time when Linux was moving up. Currently, UNIX is running on 67% of all Internet hosts. Of the Unix type OS's, 57% is running Linux and 2/3rds are running Ubuntu & Debian, and the remainder CentOS and others. That means that Linux runs 37% of all Internet servers and Windows powers 33%.
Symantec reports that "... more than 430 million new unique pieces of malware in2015, up 36 percent from the year before." Zero day vulnerabilities, discovered at more than one per week, not counting what the NSA or CIA have, is becoming a commodity. "...Vulnerabilities can appear in almost any type of software,but the most attractive to targeted attackers is softwarethat is widely used. Again and again, the majority of thesevulnerabilities are discovered in software such as InternetExplorer ...". Translation: Windows. It is, however, no secret that the vast majority of all malware is launched from Internet websites running Windows. "More than 75 percent of alllegitimate websites have unpatched vulnerabilities. Fifteenpercent of legitimate websites have vulnerabilities deemed‘critical,’ which means it takes trivial effort for cybercriminalsto gain access and manipulate these sites for theirown purposes."
With the release of Valut 7 it is only going to get worse for systems running Windows or using Windows compatibility layers. It is such layers that account of most of the vulnerabilities found in the Mac OSX.
I used to use Google cloud to store my data. For other reasons I recovered my data and deleted my Google account.
Two years ago I had only 750Gb along with two WD 350Gb USB drives, onto which I regularly dragged and dropped my home account to to backup my data. This was not entirely satisfactory because sockets, links and devices don't transfer well in a Dolphin d&d operation. It is also a pain, even when being retired and with plenty of time on my hands
, to pull out the WD USB drives and figure out which one I should drop my data on to. Sometimes I had to delete previous copies to make room. I often wondered if I had deleted something, deliberately or inadvertently, that I might want later.
I looked at software which automatically synced files to 2nd drives and/or remote locations, like a server on my LAN. An example is Syncthing. This statement didn't set well with me: "Each device scans for changes every 60 seconds...". I didn't need any more file indexing taking place.
I can't remember if I first heard about Btrfs from one of Oshunluver's posts, or from an Internet source, but the idea of COW (copy of write) beats file scanning looking for changes to backup. When the file system writes to the disk it automatically updates any snapshots, making file scanning unnecessary. A snapshot of @ or @home copies everything, files, links, sockets, devices, etc. Using send & receive to move snapshots to storage devices completed the process.
I first formatted the USB drives as Btrfs devices and mounted them as /backup. Then I'd take ro snapshots and then send them to the /backup HD. That went on for a couple years. Then my sda device started throwing sector errors so I bought two 750 Gb HDs and a CDROM caddy in which I could mount the 2nd HD. When I attempted that I discovered that a) the caddy wasn't working and b) there was a second bay in my laptop! So, I mounted the new drive as sda and the old drive as sdb and re-installed KDE Neon User Edition. Recovering my data from the 320Gb HD was easy because my entire data stash was 90Gb, Then I read a post by Oshunluver about his RAID setup and I set up my two HDs as a RAID1 pair. Perfect! Mirroring data, metadata and system data as RAID1. I got help from the ORINCO caddy support about a switch which allowed the caddy to work, so I put the other new drive into it and formatted it as Btrfs but did not add it to my pool. Strangely, when I plugged it in and turned on my laptop what was sdb became sdc and the 3rd drive became sdb. So, before adding the 3rd HD the sda was the main HD and sdb was the mirror HD. After adding the 3rd drive sda was still the main drive but sdc was now the mirror. The 3rd HD was sdb. The UUIDs kept /etc/fstab from getting confused. Now I can keep several copies of @ & @home snapshots on /backup (sdb) but most of the time it remains unmounted. If the old drive (now sdb) begins to fail I can use Btrfs procedures for replacing it without losing any data.
My Btrfs pool is now my cloud.
Symantec reports that "... more than 430 million new unique pieces of malware in2015, up 36 percent from the year before." Zero day vulnerabilities, discovered at more than one per week, not counting what the NSA or CIA have, is becoming a commodity. "...Vulnerabilities can appear in almost any type of software,but the most attractive to targeted attackers is softwarethat is widely used. Again and again, the majority of thesevulnerabilities are discovered in software such as InternetExplorer ...". Translation: Windows. It is, however, no secret that the vast majority of all malware is launched from Internet websites running Windows. "More than 75 percent of alllegitimate websites have unpatched vulnerabilities. Fifteenpercent of legitimate websites have vulnerabilities deemed‘critical,’ which means it takes trivial effort for cybercriminalsto gain access and manipulate these sites for theirown purposes."
With the release of Valut 7 it is only going to get worse for systems running Windows or using Windows compatibility layers. It is such layers that account of most of the vulnerabilities found in the Mac OSX.
I used to use Google cloud to store my data. For other reasons I recovered my data and deleted my Google account.
Two years ago I had only 750Gb along with two WD 350Gb USB drives, onto which I regularly dragged and dropped my home account to to backup my data. This was not entirely satisfactory because sockets, links and devices don't transfer well in a Dolphin d&d operation. It is also a pain, even when being retired and with plenty of time on my hands
, to pull out the WD USB drives and figure out which one I should drop my data on to. Sometimes I had to delete previous copies to make room. I often wondered if I had deleted something, deliberately or inadvertently, that I might want later. I looked at software which automatically synced files to 2nd drives and/or remote locations, like a server on my LAN. An example is Syncthing. This statement didn't set well with me: "Each device scans for changes every 60 seconds...". I didn't need any more file indexing taking place.
I can't remember if I first heard about Btrfs from one of Oshunluver's posts, or from an Internet source, but the idea of COW (copy of write) beats file scanning looking for changes to backup. When the file system writes to the disk it automatically updates any snapshots, making file scanning unnecessary. A snapshot of @ or @home copies everything, files, links, sockets, devices, etc. Using send & receive to move snapshots to storage devices completed the process.
I first formatted the USB drives as Btrfs devices and mounted them as /backup. Then I'd take ro snapshots and then send them to the /backup HD. That went on for a couple years. Then my sda device started throwing sector errors so I bought two 750 Gb HDs and a CDROM caddy in which I could mount the 2nd HD. When I attempted that I discovered that a) the caddy wasn't working and b) there was a second bay in my laptop! So, I mounted the new drive as sda and the old drive as sdb and re-installed KDE Neon User Edition. Recovering my data from the 320Gb HD was easy because my entire data stash was 90Gb, Then I read a post by Oshunluver about his RAID setup and I set up my two HDs as a RAID1 pair. Perfect! Mirroring data, metadata and system data as RAID1. I got help from the ORINCO caddy support about a switch which allowed the caddy to work, so I put the other new drive into it and formatted it as Btrfs but did not add it to my pool. Strangely, when I plugged it in and turned on my laptop what was sdb became sdc and the 3rd drive became sdb. So, before adding the 3rd HD the sda was the main HD and sdb was the mirror HD. After adding the 3rd drive sda was still the main drive but sdc was now the mirror. The 3rd HD was sdb. The UUIDs kept /etc/fstab from getting confused. Now I can keep several copies of @ & @home snapshots on /backup (sdb) but most of the time it remains unmounted. If the old drive (now sdb) begins to fail I can use Btrfs procedures for replacing it without losing any data.
My Btrfs pool is now my cloud.
Comment