Tag: security

A short while ago, Twitter got hacked. I wrote about this, as did Denny Cherry. I think Denny’s piece was more interesting, as he speculates about the security measures that Twitter may, or may not, have had in place.

One of the things Denny brings up is air-gapping administrative machines. I’ve rarely seen this in organizations, and perhaps see this less and less in the pandemic world. There are some high security places that do this, but could your organization do this? How many of us use cloud or co-location facilities where we can’t even physically enter the premises?

I suppose we can use some strong network security controls, perhaps even requiring static IP addresses for people at home and specific routes for certain administrators. I do know some companies that do require specific laptops for access, with limited software, but this certainly isn’t the norm. Too often a general laptop used for most work performs double duty as an administrative workstation with access to production data.

Another thing Denny mentions is jump hosts, without any cut/copy/paste functionality from the remote machine to the host. This is something I am starting to see from customers, even smaller ones, as a way of limiting the chance of ransomware or some security breach. Multi-factor authentication gets an administrator onto a remote desktop session on a jump host, from which they can access production systems with limited tools. This certainly isn’t perfect, and it is annoying for administrators, but it is a good security layer, and it forces organizations to use good, compliant, database DevOps practices to deploy changes.

Perhaps the best part of Denny’s article is the title of the last section: good security shouldn’t be user friendly. It shouldn’t be for administrators. While we might make things slightly easier for average users, anyone that can access bulk amounts of data, especially in a privileged fashion, should have strong security, which is a bit cumbersome. I think the hassles of strong security would be a good thing for more of us to have to deal with. Hopefully more organizations will start taking better precautions and reduce the chance of attack.

Steve Jones

Listen to the podcast at Libsyn, Stitcher or iTunes.

Imagine that you’re a researcher doing some data analysis for your company. You run some queries or load some data and then go home. The next day, you come in and find that all of our data has been deleted. Perhaps you’re the victim of a Meow attack, where people look for unsecured databases and wipe them out. If you read the comments, many of them indicate this may be considered a public service.

I think I agree with that, and here’s why. If you put up data about me in a system and don’t secure it. I’m not sure you should be trusted with the data. The article notes that UFO VPN was a victim. They got caught not only with an unsecured database, but one that had data that wasn’t supposed to be logged, including passwords. They moved their data to a new database, also unsecured, and a meow attack wiped it out.

While I understand this might cause a company to fail and affect employees who hadn’t made the decision to store this data and ignore security, I’m don’t think that the world overall is worse off because their data is gone. I’m also not sure that the employees are worse off as I’d suspect fines or other legal action might have wiped the company out anyway.

I know some university groups may lose data that is difficult or impossible to recover. I know some companies might be irreparably harmed. However, I also know that it’s 2020 and there is no reason to have an unsecured set of data available to the public. Whether a database, a file-share service, or anything else. Security needs to be provided for data.

Like many of you, I do use some services in the cloud to share files. I also find it maddening that most public access has been revoked and I need to specifically invite people, set passwords, and more to easily share things. However, that’s what we need to do in an interconnected world where we have personally identifiable and sensitive data. We need to secure it.

I’m glad SQL Server doesn’t allow blank passwords for sa, and I hope that no one allows simple, easy passwords on their systems. It is convenient, but the price your organization might pay for this convenience could put them out of business. It’s also a large price to ask someone whose data you have to pay if it impacts their life.

Steve Jones

Listen to the podcast at Libsyn, Stitcher or iTunes.

I got a new mobile phone and was going through the setup. On top of numerous Android updates, I had to reset a number of applications back up. While it’s a pain, I also appreciate from a security standpoint that moving access to my data and sites to a new device could be an issue. One of the applications I was trying to set up was Garmin Connect, as I track my exercise, heart rate, and more on my watch. When I first tried this, I got a 404 in the app, which was strange.

Apparently Garmin was having issues. They had a large outage. The next day I saw that article, and while I could get to the main Garmin site and log in, I couldn’t get a new install of the app to connect on my phone, and I couldn’t get access to any of my workout data. Even old data, apparently isn’t on my phone, as I thought. It’s being read from the cloud. That’s disconcerting, though I record my data separately at MapMyRun, so I’m not overly worried. I even found a procedure and tested it for saving my data locally.

Over the weekend after this happened, I didn’t do much, but as of the Monday after the attack, the Verge reported that some data was visible and sync was working. Not for me, but for someone. That’s good, and Garmin has a touch of information about the outage, saying they expect devices to begin syncing at some point. Mine didn’t that day, but did a day or two later.

This is interesting as an attack because it’s not just the company’s own internal data, but also their customers’ data. I don’t know if that was the intention, but what better way to put pressure on a company that take away their customers’ data. I don’t know that this would make them respond differently than losing their internal data, but it  likely would be more public. It also might put more pressure on them to pay some ransom.

Ultimately, this is an area that I think the GDPR started to help, but allowing customers to access copies of their data, as well as have rights over how it is used. I think having the right to not only get a copy of my data, but a regular backup is something I think should be required of organizations that collect information about me. Likely there might need to be a charge, but perhaps some regulation about what is reasonable and how this data should be available is a something else that might need some regulatory boundaries.

Steve Jones

Listen to the podcast at Libsyn, Stitcher or iTunes.