Usuario:Neel.arunabh/Taller
Term Definition What is a path? A path is a sequence of directories used to reference a file. They can be relative or absolute.
Why do we need file systems? 1. Locate free storage space 2. Store information in a persistent way (survives process termination) 3. Allow more than one process to access the information store concurrently 4. Find information by human-accessible identifiers 5. Protect data from unauthorized use 6. Identifying types of data stored and programs that can read the data
AD
How are directories implemented? Directories can be implemented in-line or via reference -> Variable attribute/name size
Why do we mount filesystems? We mount a file system to make it accessible to programs through the operating system.
Challenges with mounting filesystem? Where will the root of the file system appear to the user?
How should different devices be presented to the user?
When accessing a file system on a device, we need to know? boot block, Filesystem type/organization (super block), and where the root directory is
What is in the filesystem superblock? The file system superblock holds information about its type, format, and organization
What happens when you make a mounting call in Unix Copies the Superblock from the disk being mounted to the superblock table in main memory.
Sets pointers in the directory inode to FS mounted
What are the approaches to file allocation? 1. Contiguous allocation (all-in-a-row) 2. Linked list allocation (block) 3. Chained table allocation (FAT) 4. Indexed allocation (index nodes) 5. Chained indexed allocation (index + linked list) 6. Multi-level indexed allocation (tree structure)
Hardware is an abstraction that provides two main operations 1. Read block at index "k" 2. Write block at index "k" 3. Allocate and Deallocate
Contiguous allocation DEFINITION: A file's contents are stored purely in order on the drive CONS: suffers from external fragmentation, because it is a contiguous approach to system memory
Linked List Allocation DEFINITION: Each block holds a pointer to the next. Requires iteration through drive blocks (drive accesses) for random access CONS: Problematic in large files; App has to iterate through each node.
AD
Chained Table Allocation DEFINITION: Chain store the linked list as a table in memory to minimize disk access. It is called a File Allocation Table (FAT)
Indexed Allocation DEFINITION: Indexed allocation stores locations of each block in an index block referenced by the file control block CONS: A pure index system limits file size
Multi-level Indexed Allocation Use direct and indirect blocks to increase maximum file size
Chained Indexed Allocation Direct blocks with an additional indirect pointer
What are Inodes? 1. Index Nodes 2. Store file block information, along with some block pointers - but are usually smaller (128B or 256B
(T/F) Does a bigger block size have a better data rate? True
(T/F) Does a bigger block size have a good disk space utilization? False
What happens with a worse disk space utilization? Internal Fragmentation
(T/F) Does a smaller block size have a better data rate? False, lower data rate
(T/F) Does a smaller block size have a good disk space utilization? True
AD
MS-DOS File System 1. Maximum partition size for different block sizes. 2. The empty boxes represent forbidden combinations.
How do you keep track of a free block? (a) Storing the free list on a linked list (b) A bitmap
What is the file control block (FCB)? Where the attributes and location information of a file are stored. It is often distributed across the file system, meaning that they are within directories or reachable from them
Common file attributes in an FCB Owner, group, permissions, length, file type, read-lock, write-lock, date created, date modified, data accessed
True or False: In the process of creating a file, a link to the directory is created True
True of False: a deletion involves the unlinking of a file to a directory True
What is a file's link count? A counter that tracks the number of links to that file
A file's link count is __________ when a file is opened to prevent deletion mid-operation. incremented
True or False: In Linux, a file can be added to (shared by) exactly one directory via a file system link False: It can be added to more than one directory
Hard link vs soft link A hard link holds the reference to the inode of the target file (only within same file system). A soft link holds the path of the target file (may span file systems, but can suffer from link rot, which is when a file is moved and the soft link becomes invalid).
AD
What are the two types of links? Regular (hard) and symbolic (soft)
When can linking/unlinking fail? Sometimes linking/unlinking can fail due to permissions, existence, and other issues (space)
When is a file considered deleted? When its link count reaches 0
What is a file directory? A directory/folder is a special file type that holds (part of) the File Control Block about other files. A directory file is usually marked as such via an attribute in the FCB for the file
Layouts of directories Single shared directory, one directory per user, freeform/nested
What is random access? Data can be accessed in any order, and the data cursor can be moved via seek operation. Examples include HDD, SSD, Floppy)
What is sequential access? Bytes must be accessed in linear order. Some, but not all, media can rewind. Examples include tapes and pipes.
What are the two ways files can be accessed? Sequentially and random access
Methods for determining file type Extension, magic numbers (file signature), file attributes (metadata), look at file contents
What is a file? Storage that is logically named and persistent
AD
Main types of file directories FAT, VFAT, UNIX, ISO9660
What are the main categories of I/O Devices? 1) Block Devices (entire blocks; addressable (e.g. SSD)) 2) Character Devices (individual bytes; not addressable; can be one-way)
What are the types of I/O Handling Strategies? 1) Memory Mapped Input Output (MMIO) 2) Port Mapped (PMIO) 3) Direct Memory Access (DMA) 4) Dedicated Channels
Programmed I/O requires action on the part of the ___ to store and fetch from devices CPU
Human interfaces began life as _________________ Terminals
T/F: each key on a keyboard has two scan codes True
What is the difference between ASCII and Unicode? ASCII represents lower/uppercase letters, digits 0-9, and special characters. Unicode covers a larger range of characters (Greek symbols, Arabic characters, etc.)
What information is sent as mouse input Events (buttons, movement, overflow), X movement, Y movement
Types of touchscreen input 1) Resistive - measures resistance via plastic (bendable) surface 2) Capacitive - measures capacitance across glass (rigid) surface
What are the type of programmed IO? 1. Polling 2. Interrupts
AD
What is polling? Determines when a device is ready to read/write
How is a DMA (Co-processor IO) efficient? We don't need to directly handle IO, the controller does it
RAID 5 Block level striping with a distributed parity. Requires at least 3 disks. Pros: Read transactions are fast, disk failure tolerance Cons: Write transactions are improved but still low
RAID 6 Block level striping with two parity blocks allows two disk failures before any data is lost. AKA block level striping with two parity disks distributed across all member disks
Summary: Pros: High fault tolerance, storage efficiency, fast reads. Cons: Slow rebuild time (up to 24 hours)
Steps to remove a file in UNIX 1. Remove file entry from its directory entry (FCB) 2. Release the inode to the pool of free inodes 3. Return all disk blocks to free disk blocks
Journaling File Systems Write all metadata changes to a transaction log before sending any changes to disk. For any operations, the log entry is first written with the intentions, then the operations are done, and then the entry is removed. In the case of a crash, the log entry is read.
Qualities of Transmission Control Protocol (TCP) Connection oriented, lost data is re-transmitted, data is delivered in order, data is delivered as a stream, CONTROLS DATA FLOW TO MAXIMIZE EFFICIENCY
True or False: In TCP, each peer tells its counterpart how many bytes it's willing to accept (advertised window). True
True or False: In TCP, the advertised window is usually the amount of room currently available in the send buffer. False; the advertised window is usually the amount of room currently available in the RECEIVE buffer. This keeps the sender from overwhelming the receiver.
True or False: In TCP, the advertised window is constant. False; the window changes size dynamically. As data is received from the sender, the window size decreases, and as the receiver reads data from the buffer, the window size increases.
True or False: In TCP the send window is the same size as the receive window of the receiver. True
What are the two basic sections of the TCP send window? 1. Bytes that have been sent but not acknowledged 2. Bytes that can be sent (are ready or empty)
True or False: The received window moves when bytes are acknowledged (TCP) True
True or false: In TCP, when a new send window size is received from the counterpart, the send window is resized to the new window to prevent overwhelming the receiver True
What are the three steps of handling receives in TCP streams? 1. Receive the packet header 2. Receive the message's data element 3. Parse the message
RAID 4 Block-level striping with dedicated parity. Provides good performance of random reads. Random writes is low due to need to write all parity data to a single disk. AKA block level stripping with parity disk
Summary: Pros: good performance of random reads + parity. Cons: Random writes is low
RAID 3 Byte-level striping with dedicated parity. Rarely used. Generally CANNOT service multiple requests simultaneously (any single block of data is spread across all members of the set). Suitable for applications that demand the highest transfer rates in long sequential reads and writes. AKA byte level stripping with parity disk
Summary: Pros: Parity provides safety. Cons: Not practical
RAID 2 Stripes data at the bit (rather than block) level. The magnetic disks are synchronized by the controller to spin at the same angular orientation. Very HIGH TRANSFER RATES. Rarely used in practice. AKA bit stripping
Summary: Pros: High transfer rates. Cons: Not practical
Raid 0 Splits your files into blocks and scatters them across your physical storage units. It increases the overall performance (read and write speeds) due to a higher cumulative THROUGHPUT of all the storage units. Offers NO DATA REDUNDANCY and thus has no effect on data security. AKA data striping
Summary: Pros: high performance. Cons: No extra data security
RAID 1 Mirrors data between two drives withing the RAID. When data is written to one drive, it is automatically replicated in the other drive. Primary function is to provide DATA REDUNDANCY. AKA data mirroring
Summary: Pros: disk failure tolerance. Cons: Higher storage costs
What does the OS provide? 1. Buffering 2. Error reporting 3. Allocating devices 4. Uniform driver target
Why do we need a buffer in the kernel space? Having a buffer is a protection mechanism
What are the OSI Layers in order? (Bottom is on the top) 1. Physical Layer 2. Data Link Layer 3. Network Layer 4. Transport Layer 5. Session Layer 6. Presentation Layer 7. Application Layer
What does the Application Layer of OSI do? Provides the user interface
What does the Presentation Layer of OSI do? Format data (compression, encryption)
What does the Session Layer of OSI do? Opens/closes sessions between applications
What does the Transport Layer of OSI do? Transfers data reliably or unreliably. Reliable transfers involves a connection
What does the Network Layer of OSI do? Addresses packets and routes them to provide end-to-end communication [PACKET]
What does the Data Link Layer of OSI do? Packet framing. Hardware addressing (MAC) [FRAME]
What is MAC? Media Access Control address (uniquely identifies a device)
What are the layers of IP in order? (Bottom is on the top) 1. Access Layer 2. Network Layer 3. Transport Layer 4. Application Layer
What is involved in the Application Layer of IP? Application protocols
What are some examples of IP application protocols? FTP, TFTP, HTTP, SMTP, POP, Game protocols
What is involved in the transport layer of IP? TCP (Transmission Control Protocol), UDP (User Datagram Protocol)
What is is involved in the network layer of IP? IPv4, IPv6
What is the access IP layer? Ethernet and electrical connections to transmission medium
What IP layer is assisted by the OS? Application layer
What IP layers are completely handled by the OS? Access, Network, and Transport layer
What is a socket? Is one endpoint of a two-way communication link between two programs running on the network
Why are sockets important? Provides a means of inter-process communication (IPC) by establishing named contact points between which the communication takes place
What makes up a network system in IP? 1. Network interface 2. Network driver 3. Protocol stack TCP/IP
(T/F) Is IP a packet switching network protocol? True
(Y/N) Is the data sent to the network more than just your data? Yes
What are the 2 number formats that processors use? Big-Endian and Little Endian
What is Network byte order? How bytes are sent to a network (Big-Endian)
(T/F) Most processors use little-endian? True
(T/F) Can IPv6 store more addresses than IPv4? True
What is the bit-number of IPv4? 32
What is the bit-number of IPv6? 128
What are the 3 types of delivery mechanism for IP? 1. Unicast - single destination 2. Multicast - group destination 3. Broadcast - all network hosts
Why is IP better than OSI? 1. More flexible 2. Less complicated 3. Easier to implement 4. Worked from the beginning
What does the Physical Layer of OSI do? Provides electrical connections to a transmission medium [BITS]
Benefits/Drawbacks of UDP Benefits: need only one socket, faster data transmission, sends data as a whole Drawbacks: does not ensure data is sent and received in the same order, does not establish connections before sending data, no receipt of data given, guarantee that data was sent, will not retransmit/find lost data, no mechanism to handle congestion
What are some characteristics of IP? 1. Created for ARPAnet (military research) 2. Incremental implementation and model 3. Based on loosely layered model
What are some characteristics of OSI? 1. ISO Standard 2. Model before implementation 3. Based on strictly controlled layers
Why does TCP implement flow control? It prevents network congestion (can arise from sender sending data too fast for network or receiver to keep up)
T/F: The completion of a send means the data was successfully delivered False
Components of a socket address include: transport protocol, IP Address, Port Number
A port identifies the ___________ we wish to communicate with process
How does TCP handle send and receives? - send call puts information in OS buffer and sends it - receiver side will read buffer - one receive call does not correspond to one send call
What are the two types of communication protocols? 1. Open Systems Interconnection (OSI) - Good Child
2. Internet Protocol (IP) - Bad child
(T/F) Are CLOCK_REALTIME_COARSE and CLOCK_MONOTIC_COARSE faster? True
(T/F) CLOCK_REALTIME_COARSE and CLOCK_MONOTIC_COARSE are more accurate variants of CLOCK_REALTIME and CLOCK_MONOTIC? False
What is CLOCK_MONOTIC_RAW clock? 1. Gives the same as CLOCK_MONOTIC
2. Not affected by NTP adjustments
When does a counter decrement? Each pulse
What is CLOCK_MONOTONIC clock? 1. Gives the time since a fixed starting point - usually since you booted the system
2. Affected by NTP , but can't jump backward
UDP User Datagram Protocol - Connectionless method of data transfer - data arrives out of order - best effort delivery
TCP Transmission Control Protocol - Connection-Oriented Protocol, Data is delivered in-order, lost data is retransmitted
What is CLOCK_REALTIME clock? 1. Gives the time pass passed since January 1, 1970
2. Affected by NTP adjustments
What are NTP Clocks? Network Time Protocol (NTP)
Is an internet protocol used to synchronize with computer clock time sources in a network
What are the 5 types of NTP clocks? 1. CLOCK_REALTIME 2. CLOCK_MONOTONIC 3. CLOCK_MONOTONIC_RAW 4. CLOCL_REALTIME_COARSE 5. CLOCK_MONOTONIC_COARSE
What tasks need system clocks? 1. Maintaining the time of the day 2. Preventing processes from running longer than allowed 3. Accounting for CPU usage 4. Handling alarm system call from user processes 5. Providing watchdog timers for system itself 6. Profiling, monitoring, statistics
What is RAID? Redundant arrays of independent disks; creating groups of disks that are failsafe
5 Benefits of RAID 1. Provides performance gains through parallel hardware access 2. Provide reliability through redundant storage 3. Fast recovery 4. Hot swap is possible (Hot swapping is the replacement or addition of components to a computer system without stopping, shutting down, or rebooting the system; hot plugging describes the addition of components only.) 5. Can be software or hardware based
What are interrupts? Allows the device to alert the CPU that is it ready
What is failover? Failure is a backup operational mode that automatically switches to a standby database, server or network if the primary system fails
How does mirrored storage work? Every write is duplicated (write to disk 0 and disk 1). Reads are to whichever system is least loaded. Automatic failover when disk, ES, or network fails
Some display controllers support Hardware Scrolling via _____________ ______________ Pointer Adjustment
Three main types of redundancy Backups, Hardware Redundancy (mirroring, raid), Journaling (more advanced, rare)
What are the steps of Holt graph detection? 1. Eliminate "allocation only" nodes (Remove the leaf nodes) 2. As nodes eliminated, reallocate resources 3. Detect cycles from nodes via algorithm 4. Cycles are deadlock sets
What search algorithm can detect cycles? Depth First Traversal
(T/F) Do nodes appear more than once if there are cycles? True
What are the ways we can handle deadlock after we've identified it? 1. Preemption (not often possible) 2. Rollback (yuck) 3. Killing processes (ugh)
(T/F) Deadlocks can occur whenever we wait on an event? True
1. Processes may depend on one another
(T/F) We cannot run into resource deadlock in networks? False
What is Livelock? A situation in which two or more processes continuously change their states in response to changes in the other process(es) without doing any useful work.
What can cause Livelock? Start over on failure to lock (2-phase lock)
What conditions are required for livelock? Synchronization (Bad timing)
- View lecture for livelock example
What is the kernel role in IO? The user must pass data to (or request from) the kernel
What is starvation? When low-priority jobs never get run because there are always higher priority jobs running.
What are the conditions for starvation to occur? 1. Policy that makes starvation possible (some process always gets precedence over others) 2. Bad timing
- Example of starvating in lecture
What is resource contention? Refers to a conflict over a shared resource between several components
CONS: Resource contention often negatively affects the performance of components competing for the resource
What is preemptible resource? Space can be open - free resource - on RAM and restore that restore without consequences
How can deadlock be modeled? Using a Holt Graph (Watch lecture to understand this better)
What is non-preemptible resource? Cannot be release miduse. Not easy to restore. May cause resource retention
What are some common strategies to handle deadlock? 1. Ignore the problem (Yes this is true lol) 2. Detect that a deadlock has happened and recover 3. Dynamic avoidance (in realtime), (expensive) 4. Structural prevention (build without allowing deadlocks)
What are the 4 REQUIRED conditions for deadlock to occur? 1. Mutual exclusion in resource use 2. Processes may hold and wait 3. There is no resource preemption 4. There is a circular wait condition
What is the strategy to prevent deadlocks in the case of mutual exclusion? Use spooling
What are some strategies to prevent deadlocks in the case of hold and wait? 1. Denying processes holding a lock from making a request 2. Requiring all requests for locks to be simultaneous 3. Use a form of multi-resource locks 4. Use a lock server (a computer program or system that provides a central, shared mechanism for managing access to resources. It allows multiple processes or threads to request and be granted exclusive access to a resource, using mechanisms such as locks or semaphores.)
What is the strategy to prevent deadlocks in the case of no preemption? Use two-phase locking, which facilitates rollback by preventing work until all locks have been acquired
First phase (growth): Get all required locks (on checkpoints: release locks upon failure to acquire)
Second phase (shrinking): If all acquired, do work; release locks as no longer needed (or all at the end)
What is the strategy to prevent deadlocks in the case of circular wait? Use careful resource organization; order all resources and require requests to be made in order (ex: a process must acquire the resource with the lower number before the one with the higher number)
True or false: in Symmetric (Secret Key) cryptology, the decryption key is the same as (or can be easily computed from) the encryption key True
True or false: In Symmetric (Secret Key) cryptology, anyone with the key can send/store/retrieve/receive encrypted data True
What are the two main cryptographic algorithms? Symmetric (Secret Key) and Asymmetric (Public Key)
True or false: In Asymmetric cryptography, the decryption key is the same as the encryption key False; they are different, and one key cannot be easily determined from the other key
In which cryptography technique, symmetric or asymmetric, can the encryption key be announced publicly? Asymmetric
What are the three classical symmetric cryptosystems? 1. Caesar cipher: shift alphabet 2. Rail fence cipher (zig-zag): Location of letters on a text that form a sentence 3. Hill cipher: matrix multiplication, inverses, etc.)
What are the two main types of modern symmetric cryptosystems? 1. DES (digital encryption system): short key length of 56 bits makes it too insecure for modern applications 2. AES (advanced encryption system): Three different key lengths: 128 bits, 192 bits, 256 bits
True or False: asymmetric cryptography allows for the anonymity of the sender True
What is signing in cryptography? Cryptographic digital signatures use public key algorithms to provide data integrity. When you sign data with a digital signature, someone else can verify the signature, and can prove that the data originated from you and was not altered after you signed it. It is often used with a cryptographic hash/
What is the difference between authentication and identification? Identification is determining who someone is, whereas authentication is verifying a claimed identity.
Given N different characters in a password and the length of the password L, how many different passwords can there be? N^L
What is a dictionary attack? It is an attack when the perpetrator uses "more likely" passwords
What is a "Joe" account? When the password is the same as the username
What is a "finger" attack? Using knowledge of the user to guess a password or PIN
What is a loose-lipped system? When the log-in process provides information about the entered credentials. Examples include informing the user whether the username or password is invalid, rejecting a login after just the name is entered.
What is a logic bomb? A logic bomb is a string of malicious code inserted intentionally into a program to cause harm when certain conditions are met.
What are three examples of insider attacks? 1. Logic bombs 2. Back doors 3. Login spoofing
What is a back door? A usually hard coded condition that allows access to a program when certain authentication information is entered.
What is login spoofing? When a scammer copies/spoofs an existing platform to trick the user into thinking its the real thing and enter sensitive information. Ex: login page on www.amazonn.com
What is keylogging? The use of a program to record every keystroke made by a computer user in order to gain fraudulent access to passwords and other confidential information. May be used for legitimate monitoring such as biometric identification.
True or False: Auditing is not considered an important component of software defense False; it is especially important for deterring insider attacks
When does deadlock occur? 1. When a process is waiting for an event (such as release) 2. Only processes in the same set can trigger the events that another process in the set is waiting on
What is an unsafe state in dynamic avoidance? In an unsafe state, some possible futures could result in deadlock even with planning.
Three assumptions of dynamic avoidance No resources fail No process needs more than available resources Any process that gets resources it needs will complete
Four common deadlock conditions 1. Mutual Exclusion 2. Hold and wait 3. No preemption 4. Circular wait
What is a safe state in Bankers algorithm? When there is a non-deadlock sequence to process completion in the worst case. We can assess this by treating the maximum as the requested.
What are the two counts that dynamic avoidance processes have? 1. Current count for each resource type in needs 2. Maximum count for each resource type it may need
What is Bankers algorithm? The banker's algorithm is a resource allocation and deadlock avoidance algorithm. It tests for safety by SIMULATING THE ALLOCATION for predetermined MAXIMUM POSSIBLE AMOUNTS of all resources. Then it makes a "safe state" check to test for possible activities, before deciding whether allocation should be allowed to continue.
What is a safe state in dynamic avoidance? In a safe state, the resources can be allocated (to guarantee deadlock avoidance)
What are the two states possible with dynamic avoidance? Safe state an unsafe state. Dynamic avoidance attempts to avoid entering unsafe states.
What are the 3 steps in resource locking session? 1. Acquisition (Lock) 2. Utilization (Use) 3. Release (Unlock)
Why should we care about security? 1. Data loss 2. Financial loss 3. Loss of competitive edge
What does the term Dox refer to? Exposure of data
GOAL: Data confidentiality
What does the term Diddle refer to? Tampering with data
GOAL: Data Integrity
What does the term Identity theft refer to? False identification
GOAL: Authentication
What does the term unmasking refer to? Exposure identity
GOAL: Anonymity
What are some tactics and motivations that adversaries use? 1. Opportunistic prying 2. Hacktivism 3. Snooping/Espionage 4. Monetary crime 5. Direct sabotage 6. Spoofing (disguising an unknown source as being from a known trusted source)
Managing risk requires? 1. Identifying what to protect and why 2. Cost/impact of security failures and efforts 3. Security policy 4. Identifying vulnerabilities before hand 5. Mitigation measures 6. Develop and implement ad plan
What are the 3 places security can be implemented in? 1. Hardware 2. Software 3. "Wetware"
(T/F) Security does not revolve around controlling access to resources? False; it does
Give some examples of security implementation in hardware. 1. BIOS passwords 2. Encryption fobs 3. Fingerprint scanners
Give some examples of security implementation in software. 1. Bootloader 2. Operating systems 3. Applications
Give some examples of security implementation in "wetware". 1. Human beings (biometrics)
CONS: Runs the risk if failing
What is Discretionary access control? 1. Owner of object has discretion / sets access rights
2. If an agent can read it, they can share it at their discretion
What is Mandatory access control? 1. Organization sets object access rights
2. Agents do not have discretion to share - object rights are inherited (groups / labels)
What is the purpose of protection domains? Determines what access is allowed (by type of access and by object)
What is Access control matrix (ACM)? Holds information about who can access what
What happens when BIBA and Bell-Lapadula models are applied together? Single-level compartments are created
What is low-privelege? -Receiver
1. Reads system clock & compares to previous reading 2. Store 1 or 0 based on CPU usage, then yield CPU
What is high-privilege? -Sender
1. Full use of CPU quantum:1 2. Quickly yield CPU time:0
What is timing channel? -Communication via processor utilization
1. High-privilege process accesses confidential information 2. Uses CPU utilization to communicate covertly
What are some examples of Covert channel? 1. Timing channel 2. High-privilege 3. Low-privilege 4. File/Resource locking (Storage channel)
(T/F) the goals of confidentiality and integrity are often at odds? True
What is a side channel attack? Typically an attack on the physical implementation of encryption or security vs the algorithm itself. Can include collecting power consumption information, radiation, CPU sunds, and loads. Having highly resilient algorithms can help mitigate side-channel attacks.
What is a covert channel attack? A covert channel is a type of attack that creates a capability to transfer information objects between processes that are not supposed to be allowed to communicate by the computer security policy
(T/F) Agents cannot share/access information while avoiding scrutiny of system monitors? False
Rules of BIBA Model? 1. Agents at security level k can write only objects at level k or lower
2. Agents at security level k can read only objects at level k or higher
What is the BIBA Integrity model? Data integrity is the accuracy, completeness, and quality of data as it's maintained over time.
How can we address the challenges of an Access control matrix (ACM)? 1. Access Control Lists (ACLs) are by object and describe domains / access
2. Capability Lists (CLs) are by domain and describe objects / access
Rules of Bell-LaPadula Model? 1. Agents at security level k can read only objects at level k or lower
2. Agents at security level k can write only objects at level k or higher
What is the Bell-LaPadula confidentiality model? Confidentiality models are used to describe what actions must be taken to ensure the confidentiality of information. These models can specify how security tools are used to achieve the desired level of confidentiality.
What are the purpose of stack canaries? Early detection
Steps to insert to a stack canary? 1. Compiler inserts random value (canary) on stack below return address 2. When source executes return, compiler inserts code to check canary value before popping stack and jumping to return address 3. If value changed, then terminate process
What are some C/C++ limitations when it comes to security? 1. Integer overflow attacks 2. Pointer attacks 3. Buffer overflow attacks
What happens with integer overflow attacks? Can cause allocation error and denial of service
How does integer overflow occur? 1. CPU works with fixed-length numbers 2. Adding or multiplying may overflow representation 3. C programs do not catch overflow, just store and use incorrect value
How is integer overflow weaponized? 1. Feed valid but large number inputs causing overflow 2. Get process to use incorrect value for malloc
What happens with pointer attacks? Code or data is injected
What are dangling pointers? Failure to free memory that has been allocated
How can dangling pointers be weaponized? Can let others add content that is "valid" that should not actually be there
(T/F) Stack canaries are not intended to make it hard to overwrite return address on stack? False
What are some challenges of Acess control matrix (ACM)? Can get very large very quickly
1. Size is domains x objects 2. Domains include users and groups 3. Worst of all - a lot of the entries are empty!
What is Role-based access control? 1. Agents are assigned hierarchical roles (inheritable)
2. Multiple roles possible
What are the 3 approaches to access control? 1. Discretionary Access Control (DAC) 2. Mandatory Access Control (MAC) 3. Role-Based Access Control (RBAC)
(T/F) Buffer overflow attacks are a longstanding/current problem? True
What does the term DOS refer to? Denial of service
GOAL: System availability
How do buffer overflow attacks happen? If attackers know the memory layout of a program, they can intentionally feed input that the buffer cannot store, and overwrite areas that hold executable code, replacing it with their own code