Abstract
In their cloud security report last year Gartner states:
“Through 2025, more than 99% of cloud breaches will have a root cause of preventable misconfigurations or mistakes by end users.”
This statement is one of many affirmations that emphasizes the importance of data resides in the cloud.
Yet, while we migrate to cloud environments, we scrutinize the infrastructure; we secure the virtual permitter, workloads, and other assets.
We do, however, tend to forget about the security posture of our crown jowl, our data.
In this session we present, “ Snapshot”, a patent-pending technology that provides cloud data security posture.
This free agentless solution leverages cloud databases snapshot functionality and allows cloud operators to receive thorough insights of their data stored in the cloud.
To demonstrate the magnitude of the data exposure issue we will present recent research we conducted by automating the “Snapshot” technology.
The research leverages a minor users’ misconfiguration on one of the leading public cloud vendors and exposes significant number of data & sensitive records over dozens of cloud accounts.
We will present how we automate the process, what insights we accomplished from the research and reflect on our responsive disclosure process to the vendor.
Transcript
This transcript was autogenerated. To make changes, submit a PR.
Hello everyone and welcome to inside your cloud databases session.
The truth about your data security posture my name is Mor Manor
head head head head head head of innovation operations to share
with you with some insights about the importance of cloud data security posture,
as well as presenting with you one of our latest innovations called
Imperva Snapshot, a patent pending technology that can provide
robust insight about the cloud data security posture of your databases.
What we have on today at Xenja is a little bit about the current state
of cloud and data security. Then we'll jump into snapshot
technology and the importance of cloud data security posture.
Then we'll move into more technical part of the
presentation, switching gears, and we'll talk about the pattern pending technology
itself. What are the components that Imperva snapshot tool
runs in the cloud environment? And we will
wrap up the session with some real life example, the latest findings
by cloud operators and users just like you who run Imperva snapshot
tool in their environment. And what are the insights and visibility,
the type of visibility that they gain after running the
tool in their environments. Talking about cloud,
and specifically about the importance of data, I would like to
focus on those two numbers that I believe can reflect the current state
of cloud misconfiguration and data security.
If you're looking in 2020, over 31 billion
data records have been compromised,
while the majority of which were due to cloud data
breach. This is a very high numbers that emphasize to
us the magnitude of the importance to protect our
data, specifically the type of data
that resides on cloud environment. And if we look at the other number,
99%, this number form gardener's prediction
that indicates that by 2025,
99% of misconfiguration that
probably could lead to some sort of data breach would
be due to customer fault. So, meaning the majority of
those misconfiguration can be addressable had we have
the correct visibility and understanding of our cloud security posture
to put us all on the same line understanding what are the
devastating results of cloud data breach and breaches in general.
I do want to share with you a few examples from recent years. For example,
though hit the news are indicates
for small misconfiguration will leads to
devastating results due to the data breaches in those
cases. And the first is from a marketing firm that back in
June 2019 had over 340,000,000
personal records leaked due to a simple
cloud misconfiguration. The database itself was configured
to be publicly accessible and then the bad hackers put
their hands on the databases. And that leads to a very high
volume of records that have been exploited
and exposed. Another example is from a financial
firm back in July 2020 that 7
million exposed users usernames
and passwords have been exposed due to
again, could be a minor misconfiguration.
The databases was unencrypted and the s three bucket
was publicly accessible. Had a security operator had
known about those issues, I guess it would put the
necessary security rules to address those. But that
lack of visibility cloud lead to those scenarios. Data breaches
with devastating results. And with those stories in mind,
and understanding that visibility is in a key into your cloud environment,
specifically understanding the type of data that we have resides on
our databases, we've thought about what are the key
insights and visibility cloud operators needs to have on their daily
basis to understand the security posture of their cloud environment.
And those are the key pillars we're thinking
are necessary for every cloud operator. We would
like to know what is the security posture of the cloud environment
and the selected databases. If we have any vulnerabilities, are there
any software updates required? And if there are any
misconfiguration on the specific databases and the environment.
It's also really important to understand what is the type of data that we
have resides on the databases, do we have sensitive data on that?
Who have access to the database? Regarding permissions,
do we set remote access and what are the group policies
applicable on the selected database. And lastly,
in terms of compliance, it's really important to know what is the compliance
level as well as if we have any PII or PCI information on
a specific database. With that in mind,
we thought about an innovative tool that can provide visibility into
those three main pillars, misconfiguration and bad practices,
known vulnerabilities and specifically to
provide visibility into data, the cloud data security
posture indicating what is the privacy and compliance
and data classification of a
database in your environment. For that,
we came up with Imperva Snapshot, a patent pending
technology. It's important to know that at the moment,
the patent pending technology and the tool itself is dedicated
for AWS environment and RDS databases
only. And what we have in this environment is the
fact that we don't need to any credentials of the
database, the selected databases that a users want to scan.
The tool itself accesses a snapshot of a
selected databases only. We manages
to reset the master password of the snapshot
and open it on an isolated VPC.
That means the tool is production safe,
doesn't have impact on your production environment, and all the data
is also resides within the specific tenant and account
of the user. We also don't need any credential,
the ability to reset the master password and to
open the database. The snapshot database on an isolated
environment allows us to gain a lot of visibility
into the databases itself and
the data resides on the databases. When a user run
impaired a snapshot by the end of
a scan, he receives a PDF directly into your inbox
and I do want to share with you the main highlights of such a report.
First and foremost, the user resides a cloud data security posture high level
executive summary when one can learn and understand
the security posture of the selected database, it will know what
is the risk level. If there are any misconfiguration of the databases,
what are the main vulnerabilities of the database as well as the sensitive data
found on the selected database. A users can also see
what are the top insights and main acts for mitigation
in order to have a solid security posture on the selected database.
As we go through the report, we'll have a deep dive section for each
one of the categories for permissive configuration and best practices.
A user can see what are the security risks, what are the assessment tests
that have been run and what are the outcome of each one of those.
And obviously all of these are being listed based on the severity
level. In terms of known vulnerabilities. The user can
see what are the type of vulnerabilities that have been found that the database
is exposed to being labeled by years later the year.
Obviously it requires a patch. Probably that CV
is being exploited and exploited somewhere in the wild, but the user
really can understand what is the current states of the
database and what is the security level of the selected database that has been
scanned. And lastly, in terms of
cloud data security posture,
imperfect snapshot provide visibility into data resides on the
databases, making the labeling of the sensitive data records that have been
found. What are the categories of each, what are the total items and
on which column and what overall number of columns are
including sensitive data.
The user can also see the compliance level for the selected
databases. Understanding what is the compliance degree for
the selected scan. Understanding what the tool is
doing is great. You get an interesting insight and
a quick posture of the selected databases as
well as visibility into the data.
But I do want to take you through the journey of how the magic happens
and what happened behind the scene. When you're running parava snapshot,
it's important to know that the entire process is fully automated.
As mentioned previously, currently the tool supports
AWS environment and RDS and
Aurora databases, and I do want to take you through the phases
of what is being triggered and what tool that we use
when running parabas snapshot. So really starting that
automated process with basic prerequisite,
obviously we need to make sure that the selected database
is RDS or Aurora and we need to make sure that
the specific RDS has an available snapshot.
Without the snapshot the scan cannot be completed. This specific
snapshot is then being deployed in an isolated VPC and for
that we're also checking that there is an available
VPC to deploy that snapshot into.
Once those prerequisites are done and being checked, I do want to take you
through the journey of provisioning in flow and what are the basic templates that
we use on the cloud formation. It's important to know that
in this point all templates are publicly available,
so imperva snapshot tool is free and production safe
and we have full transparency to help our user to understand
what is being run in their environment while running imperfect snapshot.
So the root templates actually start by collecting all the key
parameters and is the base template to the other templates
that are being used. The setup template and the installer template
the setup templates start with verifying the prerequisite from the slide before
and making sure that the selected database and the DB
identifier are on the same region. Then moving forward into
the installer template where all lambdas
are being set to run pervasnapso tool
itself and to start assess the database
and route all the relevant tests. So what lambdas
are being created and what are the functions of lambda in this process?
The first and foremost is the sandbox or lambda that acts
the devsec expert in the environment. We can say
it creates all the networking resources, it creates the next
lambdas, it restore the selected databases into
the isolated vpcs and it delete all created
resources by the end of the process.
So we make sure that everything is
running on an isolated environment that will have no
impact on your live production environment.
And by the end of the process the isolated VPC as
well as other lambdas and templates are automatically deleting themselves
that they will have no impact on your environment.
The next lambda is the scanner and the key
function of the scanner is pierced to reset the master
password of the restored RDS. This is one of the main part
of the patent pending of Imperibus snapshot, the fact that you don't need
to insert database credentials at any time.
The scanner Lambda delete its master
password and then once you deploy it on the isolated VPC,
we have full access to the databases on the
isolated environment to run all queries and assessment
tests. It ran the scan as well as
classification, and it sends all the data into the
last lambda, the reporter that collects all the results from
the scanners and also generate a PDF report that is
being sent to the user. So by the end of that process,
a user who runs Imperva snapshot in their environment
received that Imperva
snapshot PDF report to gain visibility
into the cloud data security posture of the selected database.
Understanding why the components are great, but I do want to take with you
a step by step through the flow and what is the
process and what's happening through an Imperva
snapshot scan. And if we start with the installer stack who
creates the sandbox lambda, then it creates
the isolated VPC where we can deploy
the selected snapshot.
Obviously it's based on the fact that it
all matched the prerequisite and the selected database has
an available snapshot to be deployed.
Once the snapshot is being deployed and
being restored into that isolated VPC, we create a
security group from the sandbox
or lambda in order to make sure that there is
no Internet gateway and all assets have capabilities
and yam rules only to do relevant
activities and functionalities to impair the snapshot. So the
entire environment is isolated and each one of the resources and
the Yam rule created during the process have access and
permissions to do activities only on the isolated
environment and on the isolated components.
Then there scanner Lambda start dire assessment
and report all insight into the reporter who then
send that final report to the users
with the impairment snapshot final report I mentioned through
the process a few times the importance of production safe
tool obviously for cloud operators to run a new tool in
their environment, that might be tricky
at times, but we do have full transparency of
what Imperva snapshot is doing and we did want to make sure that this is
a really production safe tool. So as mentioned before, we have a
full transparency of imperfect snapshot and it's code and
it's available on our GitHub project. But specifically regarding
the precautions that we have taken developing the tool,
we made sure that the snapshot
is being restored on an isolated VPC. New security
groups in the routing table are being created and no Internet gateway
is in place, so nothing can come in and out from
that isolated VPC. So also important to
know that only resources that were created during that
process the new security groups have access into so
it cannot have access into other resources previously,
resources that have been created in the environment.
Another important note as part of the patent pending
technology of impairment snapshot is the fact that we don't need
an admin permissions to run the tool. Security admin
or cloud tech operators who want to know the security posture of
the selected database can simply run the tool.
No need to admin permissions and no need for credential
for the selected databases. This is a real
game changer that can make your life easier understanding the current security posture
of a selected databases for the
final part of the presentation, I do want to share with you some examples
based on recent scanning and findings from
Imperva snapshot scans by cloud operators just like you
who run Imperva snapshot in their environment and
share with you some insights, common insights that have been identified against
all those users who came up to us seeking more information
about impaired snapshot and how to mitigate the security gaps
in their environment. So the first example
referred to into a marketing organization that had a few
misconfiguration in their environment. Imperva Snapshot has identified
that there were a few misconfiguration about excessive
permissions. The database itself was not encrypted
and a snapshot of the databases was publicly available.
Think about it to yourself for a second. So a bad act has
access to a publicly available snapshot and the database itself is
not even encrypted. Easy win for the bad guys.
Specifically, the databases was also with over 200 known
vulnerabilities, so this is a really bad shape in terms of security
posture for the selected database. Now, if we make
the correlation in order to understand what would be the ideal game plan
and what we need to mitigate, then we have the added
value of cloud data. Security posture and Imperva
snapshot indeed indicate what is the type of data and sensitive
data that have been found on the selected
databases. So over 1.2 million email
addresses have been stored and resides on the database. Over five
hundred k of personal identifiers and over 100,000
mobile phone numbers have been identified
in that scan. So within those
misconfiguration and with that type of sensitive data resides
on the database that could really have been and become
another one of those base scenarios of a massive data
leakage. Another scenario from a financial
firm also points out a common misconfiguration that
we see in an example before databases that was not encrypted,
a snapshot was publicly available.
We also have seen that the specific databases
had the subscription disabled for security group updates,
so you don't get alerts of any security groups for database
specifically, we have found impermanent snapshot has
identified 500 passwords,
100 credit cards and almost 2000 mobile
numbers. Again, a high amount of sensitive
data within those misconfiguration in place can be easily
leaked into the bad guy. And the
last example that I want to bring you is from a cloud service company
that also, as you can see, there is some sort of a
repeating motive. The type of misconfiguration,
a misconfiguration that lead into
excessive permissions, a database that was not encrypted, and a publicly
available snapshot, the specific database,
have been vulnerable to almost 100 known vulnerabilities.
And look about the overall sensitive data that have been stored there.
20,000 addresses, 20,000 email addresses
and close to five k of personal data.
Make the correlation between the security states
of the databases, the misconfiguration that you have, and the sensitive
data being reside on the databases.
A cloud operator can come with a game plan and decide what
database is the top priority to be addressed and
protect and address the security gaps that have been found.
And this is the key thing by Imperva Snapshot,
making that correlation providing that visibility between
misconfiguration, understanding the security posture of
the databases, and what is the type of
data that is being resides on the databases.
And based on that, you can come up with a game plan
to address the security gaps and with
priority and visibility to the content and data
that you want to protect. So in terms of key takeaways
and the common misconfiguration, and what we have seen in the wild from
users in cloud operator, again, just like you who run cloud environment,
we saw that there are some misconfiguration that repeat
themselves and users need to be careful and pay
attention to address those ahead of mind so they will not be
part of the statistics. Going back to that Gartner quoted in the
beginning of the presentation, making sure you
don't have any misconfiguration when it comes into excessive permissions,
publicly available snapshot, it's really easy to fall into this
one. But even if you have a lab environment and you want to
share the snapshot between two different location, and you
want to have access between two different locations,
make sure that if you make a snapshot publicly
available for a certain amount of time, change it into a private mode
right away. Otherwise it's being available out there.
And if your data is not encrypted, the data on
that specific snapshot or on the database is not encrypted.
That's really an easy game and an easy win for the Badax.
And finally, in terms of subscription for security groups,
make sure it's being enabled to be triggered regarding
any security alerts that are related
to your databases.
I really hope you find Imperva Snapshot interesting, and if you want to
learn more to take some of that transparency
to see the code of Imperva Snapshot, I first welcome you to
look into our GitHub project where you can see also our
CLI project in our GitHub where
Imperva snapshot can be activated through a CLI where you can
schedule tasks, automate the process, and get
an immediate posture security posture of a new database that has
been created. And I also recommend you to try impervisnapt yourself if
you go into our landing page. Try Imperva.com
slash snapshot and see it yourself. Get some more information
if you want, run it on your environment and get
some insights to make sure your cloud databases
is fully secure. If you want to reach out, feel free.
Mor Manor hit me up the email or on LinkedIn.
I'm really happy to share those insights with you and hopefully talk
with you soon. Enjoy the rest of the day.