Replies

I am aware of Macrium’s guidance to unmount file based virtual disks before imaging a disk that contains the encrypted file. That said, I have never encountered a problem restoring the contents of a virtual disk even if it was mounted when the host disk was being imaged (which is often the case). I understand the risk, especially if the contents of the virtual disk are under going rapid change during the backup. My encrypted virtual disks are relatively small (1-2 GB) which limits the exposure window as they are backed up quickly.

How does Onedrive/win8 backup system work with Outlook.pst files? I have found other cloud services not to deal with them. Too large a file, and constantly changing.

Copying active .pst files to the ‘cloud’ is a bear in my experience. Many backup tools can’t handle open files and those that do tend to see the file as a constantly changing blob which yields huge incremental uploads. I use JungleDisk for my business clients as it is one of the few services that utilizes ‘data deduplication’ which dramatically reduces online storage requirements. Unfortunately, JungleDisk hasn’t been updated for several years and only supports Amazon S3 and ‘Cloud Files’ which is the default storage pool and a pricey alternative to S3.

I don’t depend on JungleDisk (JD) for most restores; I much prefer system images as noted in a previous post. However, JD provides a reliable and automated method for backing up hefty Outlook .pst files to the cloud. I’m sure there are a few other alternatives; perhaps someone else can post their experiences.

Here’s a bit more on data deduplication for the technically curious: http://en.wikipedia.org/wiki/Data_deduplication

As usual Fred’s article is well written, well intended and great advise to someone looking for a native backup solution. Thank you! Unfortunately, my experience with File History has been less than satisfactory with unexplained failures occurring after weeks/months of normal operation. Manifestations include files/directories that fail to get backed up and zombie processes that consume resources but perform no useful work. This is more prevalent when more than one user account is configured to use File History. Pretty big let down when you go to recover a file and find out File History has been off-line for weeks with no notification. The most serious failure modes apparently survive reboots with no overt notification. I suspect something gets written to the Windows Event log but it certainly isn’t prolific.

My solution is to image machines weekly with daily incremental backups. Storage requirements are considerably higher than File History but a high capacity external (USB) storage is dirt cheap. I happen to use Macrium Reflect but any credible imaging solution will get the job done. On occasion I copy a full image to another external drive and store it off-site (sneaker net). I also copy critical personal files to cloud storage at various intervals using several different methods. Most happen automatically but a few I need to kick off manually. Really depends on the application, data importance and whether pre-encryption is needed before copying to the cloud.

Over many years I have had the need to recover individual files, full directories and entire machines. The method described above has never let me down.

I recognize Fred’s solution is fully automated as would be preferable by many users if it was reliable. My experience on several different personal and client machines has proven otherwise. Once bitten, twice shy …

Simply Turn OFF Google’s Location Tracking. It takes one second. [ ] Enable Location Sharing

The OP Woody discussed Google Location a lot in the article but didn’t bother to mention how easy it is to turn off. All gloom and doom. Boo. Not.

Woody’s story is relevant but also reeks of fear mongering. Google makes it very easy to view and change Location Reporting and Location History. Google also provides a means to delete previously collected history (I generally trust they will delete info as requested but obviously it could be retained potentially resurrected in another form). Yes, both are enabled by default which I wish wasn’t the case but Google is a business and location based services have proven both popular and profitable. I’m no Google fanboy but do understand the business model.

I expected more from an experienced tech journalist. Woody should know about Google’s location based services on Android or at least should have done a bit more research before publishing this article. The point he is trying to drive (improved privacy legislation) is muted by unnecessary scare tactics.

I am with Millwood and JohnReam on this one. First thing that came to mine is *how* a brute force attack could be successful against a site that promotes strong security practices. No eat own dog food?

I have been using MSE on upwards of a dozen machines (some augmented by Threadfire or Prevx) since 4Q09 with no issues. While full scans can take a long time they do not place an excessive burden on the system and resources are easily relinquished to other applications as needed. Also, no mention was made of quick scans which we have configured to run daily and typically complete within a few minutes. After an initial full scan I tend to stick with quick scans as real time monitoring and other sensible interventions should catch most nasties – at least in the environments/systems that I support. MSE goes in easy and stays out of the way which are key criteria for my clients. Uninstall is also quick and clean should the need arise.