1 Chapter 11: Storage and File StructureRevised in March 2010

2 Chapter 11: Storage and File StructureOverview of Physical Storage Media Magnetic Disks RAID Tertiary Storage Storage Access File Organization Organization of Records in Files Data-Dictionary Storage

3 Classification of Physical Storage MediaSpeed with which data can be accessed Cost per unit of data Reliability data loss on power failure or system crash physical failure of the storage device Can differentiate storage into: volatile storage: loses contents when power is switched off non-volatile storage: Contents persist even when power is switched off. Includes secondary and tertiary storage, as well as battery-backed up main-memory.

4 Physical Storage MediaCache – fastest and most costly form of storage; volatile; managed by the computer system hardware (e.g. CPU) Note: “Cache” is pronounced as “cash” Main memory: fast access (10s to 100s of nanoseconds; 1 nanosecond = 10–9 seconds) generally too small (or too expensive) to store the entire database capacities of up to a few Gigabytes widely used currently Capacities have gone up and per-byte costs have decreased steadily and rapidly (roughly factor of 2 every 2 to 3 years) Volatile — contents of main memory are usually lost if a power failure or system crash occurs.

5 Physical Storage Media (Cont.)Flash memory Data survives power failure Reads are roughly as fast as main memory Less than 100 nanoseconds Data can be written at a location only once, but location can be erased and written to again No in-place update !!! Can support only a limited number (10K – 1M) of write/erase cycles. Erasing of memory has to be done to an entire bank of memory But writes are slow (few microseconds), erase is slower

6 Physical Storage Media (Cont.)Flash memory A form of EEPROM (electrically erasable programmable read-only memory) NOR Flash Fast reads, very slow erase, lower capacity Used to store program code in many embedded devices NAND Flash Page-at-a-time read/write, multi-page erase High capacity (several GB) Widely used as data storage mechanism in portable devices USB (universal serial bus) memory stick

7 Physical Storage Media (Cont.)Magnetic-disk Data is stored on spinning disk, and read/written magnetically Primary medium for the long-term storage of data; typically stores entire database. Data must be moved from disk to main memory for access, and written back for storage direct-access – possible to read data on disk in any order, unlike magnetic tape Survives power failures and system crashes disk failure can destroy data: is rare but does happen

8 Physical Storage Media (Cont.)Optical storage non-volatile, data is read optically from a spinning disk using a laser CD-ROM (640 MB) and DVD (4.7 to 17 GB) most popular forms Write-one, read-many (WORM) optical disks used for archival storage (CD-R, DVD-R, DVD+R) Multiple write versions also available (CD-RW, DVD-RW, DVD+RW, and DVD-RAM) Reads and writes are slower than with magnetic disk Juke-box systems, with large numbers of removable disks, a few drives, and a mechanism for automatic loading/unloading of disks available for storing large volumes of data

9 Difference bet. “-” and “+”First of all, you need to identify which format your DVD Player able to support. This can be easily know by finding the logo sticker either at the front or the back of your DVD Player. (usually DVD-R,DVD-RW,DVD+R,DVD+RW,DVD-RAM). Basically "-R" or "+R" is stand for "RECORDABLE"; while "-RW" or "+RW" is stand for "Re-Writeable". Just remember that the "-" MINUS and "+" PLUS sign is only act as an identifier of the two different format of the DVD used as they are not compatible to each other. Meaning a DVD-R/DVD-RW Writer can only writes a DVD-R/DVD-RW disc while DVD+R/DVD+RW Writer can only writes a DVD+R/DVD+RW disc(unless you are using a Combo drive that writes both formats). DVD "+" is supported by Philips, Sony, Hewlett-Packard, Dell, Ricoh, Yamaha, and others. Even though it is not supported by DVD Forum, it still gain its popularity and co-exist with DVD "-" in the market. "R" here means the disc is only allowed you to write once. Once the data is burned, it will stays there permanently. In other words, it cannot be erase. However, if you only burned, say like 1GB of data to the DVD-R or DVD+R, you will still have 3GB space available to burn next time. The system will append your new burned data instead of replacing the old data. "RW" here means the data burned before can be erased and replace by new data. Data appending is always allowed-able as long as there is enough space left on the DVD disc.

10 Physical Storage Media (Cont.)Tape storage non-volatile, used primarily for backup (to recover from disk failure), and for archival data sequential-access – much slower than disk very high capacity (40 to 300 GB tapes available) tape can be removed from drive  storage costs much cheaper than disk, but drives are expensive Tape jukeboxes available for storing massive amounts of data hundreds of terabytes (1 terabyte = 109 bytes) to even a petabyte (1 petabyte = 1012 bytes)

11 Storage Hierarchy

12 Storage Hierarchy (Cont.)primary storage: Fastest media but volatile (cache, main memory). secondary storage: next level in hierarchy, non-volatile, moderately fast access time also called on-line storage E.g. flash memory, magnetic disks tertiary storage: lowest level in hierarchy, non-volatile, slow access time also called off-line storage E.g. magnetic tape, optical storage

13 Magnetic Hard Disk MechanismNOTE: Diagram is schematic, and simplifies the structure of actual disk drives

14 Magnetic Disks Read-write headPositioned very close to the platter surface (almost touching it) Reads or writes magnetically encoded information. Surface of platter divided into circular tracks Over 50K-100K tracks per platter on typical hard disks Each track is divided into sectors. Sector size typically 512 bytes Typical sectors per track: 500 (on inner tracks) to 1000 (on outer tracks) To read/write a sector disk arm swings to position head on right track platter spins continually; data is read/written as sector passes under head

15 Magnetic Disks (Cont.) Head-disk assembliesmultiple disk platters on a single spindle (1 to 5 usually) one head per platter, mounted on a common arm. Cylinder i consists of ith track of all the platters Earlier generation disks were susceptible to “head-crashes” leading to loss of all data on disk Current generation disks are less susceptible to such disastrous failures, but individual sectors may get corrupted

16 Disk Controller Disk controller – interfaces between the computer system and the disk drive hardware. accepts high-level commands to read or write a sector initiates actions such as moving the disk arm to the right track and actually reading or writing the data Computes and attaches checksums to each sector to verify that data is read back correctly If data is corrupted, with very high probability stored checksum won’t match recomputed checksum Ensures successful writing by reading back sector after writing it Performs remapping of bad sectors

17 Disk Subsystem

18 Storage Interfaces Disk interface standards familiesATA (AT attachment) range of standards Evolved from Western Digital’s IDE (integrated drive electronics) interface PATA (parallel ATA) : only allows cable length up to 18 inches Largely replaced by SATA since 2007 SATA (Serial ATA) SCSI (Small Computer System Interconnect) range of standards Several variants of each standard (different speeds and capabilities) Fiber Channel interface for large systems (e.g. SAN) FireWire interface (apple, external disk systems)

19 Storage area network (SAN)A specialized, high-speed network attaching servers and storage devices, whose primary purposes is the transfer of data between computer systems and storage elements (such as disk arrays, tape libraries, and optical jukeboxes) An architecture to attach remote storage devices to servers in a way that the devices appears as locally attached to the operating system Disks can be shared by multiple computing servers Allows “any-to-any” connection across the network, using interconnect elements such as routers, gateways, hubs, and switches A SAN alone does not provide the “file” abstraction, only block-level operations !!! Cost and complexity of SANs dropped in the late 2000’s, resulting in wider adoption

20 SAN: The network behind the servers

21 Network-attached storage (NAS)NAS is file-level computer data storage connected to a computer network, providing data across to heterogeneous network clients The key difference between direct attached storage (DAS) and NAS is that DAS is simply an extension to the an existing server and is not networked, while NAS sits on a network as its own entity; it is easier to share files with NAS NAS uses file-based protocols such as NFS (popular on UNIX systems), SMB/CIFS (Server Message Block/Common Internet File System, used with MS Windows), AFP (Apple Filing Protocol, used with Apple Macintosh computers), or AFS (Andrew File System, developed by Carnegie Mellon University) Note, SAN Protocols are SCSI, Fiber Channel, iSCSI, ATA over Ethernet (AoE), or HyerSCSI There is SAN-NAS hybrid, offering both file-level protocols (NAS) and block-level protocols (SAN) from the same system. Openfiler, a free software product running on Linux

22 Performance Measures of DisksAccess time – the time it takes from when a read or write request is issued to when data transfer begins. Consists of: Seek time – time it takes to reposition the arm over the correct track. Average seek time is 1/2 the worst case seek time. Would be 1/3 if all tracks had the same number of sectors, and we ignore the time to start and stop arm movement 4 to 10 milliseconds on typical disks Rotational latency – time it takes for the sector to be accessed to appear under the head. Average latency is 1/2 of the worst case latency. 4 to 11 milliseconds on typical disks (5400 to r.p.m.) 8 to 20 milliseconds Compare them with main-memory access time !!!

23 Performance Measures (Cont.)Data-transfer rate – the rate at which data can be retrieved from or stored to the disk. Up to hundreds MB per second lower for inner tracks (since they have fewer sectors) Multiple disks may share a controller, so rate that controller can handle is also important ATA-5: 66 MB/sec PATA: 133 MB/sec SATA: 150 MB/sec Ultra 320 SCSI: 320 MB/s Fiber Channel (FC2Gb): 256 MB/s

24 Performance Measures (Cont.)Mean time to failure (MTTF) – the average time the disk is expected to run continuously without any failure. Probability of failure of new disks is quite low, corresponding to a theoretical MTTF of 500,000 to 1,200,000 hours (57 to 136 years) for a new disk E.g., an MTTF of 1,200,000 hours for a new disk means that given 1000 relatively new disks, on an average one will fail every 1200 hours MTTF decreases as disk ages !!! Most disks have an expected life span of 3 to 5 years !!!

25 Optimization of Disk-Block AccessWhy we have to consider this issue??? Access on disk is several orders of magnitude slower than access to main memory !!! This is an unavoidable fact, which is never likely to change in the future Block – a contiguous sequence of sectors from a single track data is transferred between disk and main memory in blocks Typical block sizes today range from 4 to 16 kilobytes

26 Schedule Issues Disk-arm-scheduling algorithms order pending accesses to tracks so that disk arm movement is minimized elevator algorithm : move disk arm in one direction (from outer to inner tracks or vice versa), processing next request in that direction, till no more requests in that direction, then reverse direction and repeat

27 File Organization IssuesFile organization – optimize block access time by organizing the blocks to correspond to how data will be accessed Store related information on the same or nearby blocks/cylinders. File systems attempt to allocate contiguous chunks of blocks (e.g. 8 or 16 blocks) to a file Files may get fragmented over time E.g. if data is inserted to/deleted from the file Or free blocks on disk are scattered, and newly created file has its blocks scattered over the disk Sequential access to a fragmented file results in increased disk arm movement Some systems have utilities to defragment the file system, in order to speed up file access

28 Non-volatile RAM (NV-RAM)Nonvolatile write buffers speed up disk writes by writing blocks to a non-volatile RAM buffer immediately Non-volatile RAM: battery backed-up RAM or flash memory Even if power fails, the data is safe and will be written to disk when power returns Controller then writes to disk whenever the disk has no other requests or request has been pending for some time Database operations that require data to be safely stored before continuing can continue without waiting for data to be written to disk Writes can be reordered to minimize disk arm movement NV-RAM buffers are found in modern RAIDs

29 Log Disk Journal is a disk area devoted to writing a sequential log of block updates Write to log disk is very fast since no seeks are required Usually writes to journal internal file information (metadata such as file allocation information), not data logs generated by user applications At periodic internals, the journal is committed to the file system Used exactly like non-volatile RAM Journaling file system keeps track of the changes it intends to make in a journal (usually a circular log in a dedicated area of the file system). In the event of a system crash or power failure, such file systems are quick to bring back online and less likely to become corrupted Today, several journaling file systems are actively used JFS2 (IBM AIX) XFS (by Silicon Graphics, now ported to Linux) Third extended file system (ext3fs) Note, no need for special hardware (NV-RAM)

30 RAID RAID: Redundant Arrays of Independent (or inexpensive) Disksdisk organization techniques that manage a large numbers of disks, providing a view of a single disk of high capacity and high speed by using multiple disks in parallel, and high reliability by storing data redundantly, so that data can be recovered even if a disk fails The chance that some disk out of a set of N disks will fail is much higher than the chance that a specific single disk will fail. For example, a system with 100 disks, each with MTTF of 100,000 hours (approx. 11 years), will have a system MTTF of 1000 hours (approx. 41 days)

31 Improvement of Reliability via RedundancyRedundancy – store extra information that can be used to rebuild information lost in a disk failure Mirroring (or shadowing) Duplicate every disk. A logical disk consists of two physical disks. Every write is carried out on both disks Reads can take place from either disk If one disk in a pair fails, data still available in the other Data loss would occur only if a disk fails, and its mirror disk also fails before the system is repaired Probability of combined event is very small Except for dependent failure modes such as fire or building collapse or electrical power surges

32 Improvement of Reliability via RedundancyMean time to data loss depends on mean time to failure, and mean time to repair For example, MTTF of 100,000 hours, mean time to repair of 10 hours gives mean time to data loss of 500*106 hours (or 57,000 years) for a mirrored pair of disks (ignoring dependent failure modes) In practice, the assumption of independence of disk failures is not valid. Power failures or natural disasters (earthquake, fire, flood) happen.

33 Improvement in Performance via ParallelismTwo main goals of parallelism in a disk system: 1. Load-balance multiple small accesses (block accesses) to increase throughput 2. Parallelize large accesses to reduce response time. Improve transfer rate by striping data across multiple disks. Bit-level striping – split the bits of each byte across multiple disks But seek/access time worse than for a single disk Bit level striping is not used much any more Block-level striping – with n disks, block i of a file goes to disk (i mod n) + 1 Requests for different blocks can run in parallel if the blocks reside on different disks A request for a long sequence of blocks can utilize all disks in parallel

34 RAID Levels RAID organizations, or RAID levels, have differing cost, performance and reliability characteristics RAID Level 0: Block striping; non-redundant. Used in high-performance applications where data lost is not critical. RAID Level 1: Mirrored disks with block striping Offers best write performance. Popular for applications such as storing log files in a database system.

35 RAID Levels (Cont.) RAID Level 2: Memory-Style Error-Correcting-Codes (ECC) with bit striping.

36 RAID Levels (Cont.) RAID Level 3: Bit-Interleaved Paritya single parity bit is enough for error correction, not just detection When writing data, corresponding parity bits must also be computed and written to a parity bit disk To recover data in a damaged disk, compute XOR of bits from other disks (including parity bit disk) Faster data transfer than with a single disk, but fewer I/Os per second since every disk has to participate in every I/O.

37 RAID Levels (Cont.) RAID Level 4: Block-Interleaved Parity; uses block-level striping, and keeps a parity block on a separate disk for corresponding blocks from N other disks. When writing data block, corresponding block of parity bits must also be computed and written to parity disk To find value of a damaged block, compute XOR of bits from corresponding blocks (including parity block) from other disks.

38 RAID Levels (Cont.) RAID Level 4 (Cont.)Provides higher I/O rates for independent block reads than Level 3 block read goes to a single disk, so blocks stored on different disks can be read in parallel Before writing a block, parity data must be computed Can be done by using old parity block, old value of current block and new value of current block (2 block reads + 2 block writes) Or by recomputing the parity value using the new values of blocks corresponding to the parity block More efficient for writing large amounts of data sequentially Parity block becomes a bottleneck for independent block writes since every block write also writes to parity disk

39 RAID Levels (Cont.) RAID Level 5: Block-Interleaved Distributed Parity; partitions data and parity among all N + 1 disks, rather than storing data in N disks and parity in 1 disk. E.g., with 5 disks, parity block for nth set of blocks is stored on disk (n mod 5) + 1, with the data blocks stored on the other 4 disks.

40 RAID Levels (Cont.) RAID Level 5 (Cont.)Higher I/O rates than Level 4. Block writes occur in parallel if the blocks and their parity blocks are on different disks. Subsumes Level 4: provides same benefits, but avoids bottleneck of parity disk.

41 RAID Levels (Cont.) RAID Level 6: P+Q Redundancy scheme; similar to Level 5, but stores extra redundant information to guard against multiple disk failures. Better reliability than Level 5 at a higher cost; not used as widely.

42 RAID Levels (Fig. 11.3)

43 Choice of RAID Level (1/2)Factors in choosing RAID level Monetary cost Performance: Number of I/O operations per second, and bandwidth during normal operation Performance during failure Performance during rebuild of failed disk Including time taken to rebuild failed disk RAID 0 is used only when data safety is not important E.g. data can be recovered quickly from other sources Level 2 and 4 never used since they are subsumed by 3 and 5 Level 3 is not used since bit-striping forces single block reads to access all disks, wasting disk arm movement Level 6 is rarely used since levels 1 and 5 offer adequate safety for most applications So competition is mainly between 1 and 5

44 Choice of RAID Level (2/2)Level 1 provides much better write performance than level 5 Level 5 requires at least 2 block reads and 2 block writes to write a single block, whereas Level 1 only requires 2 block writes Level 1 preferred for high update environments such as log disks Level 1 had higher storage cost than level 5 disk drive capacities increasing rapidly (50%/year) whereas disk access times have decreased much less (x 3 in 10 years) I/O requirements have increased greatly, e.g. for Web servers When enough disks have been bought to satisfy required rate of I/O, they often have spare storage capacity so there is often no extra monetary cost for Level 1! Level 5 is preferred for applications with low update rate, and large amounts of data Level 1 is preferred for all other applications

45 Hardware Issues Software RAID: RAID implementations done entirely in software, with no special hardware support Hardware RAID: RAID implementations with special hardware Use non-volatile RAM to record writes that are being executed Beware: power failure during write can result in corrupted disk E.g. failure after writing one block but before writing the second in a mirrored system Such corrupted data must be detected when power is restored Recovery from corruption is similar to recovery from failed disk NV-RAM helps to efficiently detected potentially corrupted blocks Otherwise all blocks of disk must be read and compared with mirror/parity block

46 Hardware Issues (Cont.)Hot swapping: replacement of disk while system is running, without power down Supported by some hardware RAID systems, reduces time to recovery, and improves availability greatly Many systems maintain spare disks which are kept online, and used as replacements for failed disks immediately on detection of failure Reduces time to recovery greatly Many hardware RAID systems ensure that a single point of failure will not stop the functioning of the system by using Redundant power supplies with battery backup Multiple controllers and multiple interconnections to guard against controller/interconnection failures

47 RAID Terminology in the IndustryRAID terminology not very standard in the industry E.g. Many vendors use RAID 1: for mirroring without striping RAID 10 or RAID 1+0: for mirroring with striping “Hardware RAID” implementations often just offload RAID processing onto a separate subsystem, but don’t offer NVRAM. Read the specs carefully! Software RAID supported directly in most operating systems today

48 RAID Example: EMC CX4-120 (1/2)Components A modular 2 Gb/s storage processor enclosure A 4 Gb/s UltraPoint™ disk-array enclosure Dual standby power supplies Features Support for both direct-attach and SAN environments Hot-pluggable I/O modules- 4 Gb/s FC, 1 Gb/s iSCSI Hot-swappable storage processors with up to 3GB of memory per SP RAID level 0, 1, 1/0, 3, 5, and 6, individual disk support, and global hot sparing Five-drive minimum to 120-drive maximum system configuration

49 RAID Example: EMC CX4-120 (2/2)Classification Specification Architecture UltraScale Controllers 2EA Processor Dual Core 1.2G Xeon *2EA Physical memory 6GB Drivers 5~120EA Fibre Channel ports 4~12EA (4G) iSCSI ports 4~8EA (1G) Disk interface 2EA (4G) RAID level 0, 1, 1/0, 3, 5, 6

50 Optical Disks Compact disk-read only memory (CD-ROM)A fairly large capacity (640M), cheap Seek time about 100 msec (optical read head is heavier and slower) Higher latency (3000 RPM) and lower data-transfer rates (3-6 MB/s) compared to magnetic disks Digital Video Disk (DVD) DVD-5 format holds 4.7 GB , variants up to 17 GB Slow seek time, for same reasons as CD-ROM Recently, HD-HVD (12-13 GB), blu-ray DVD (25-50 GB) Record once versions (CD-R and DVD-R) Jukeboxes – devices that store a number of optical disks (up to several hundreds) and load them automatically on demand to a number of drives (usually 1-10)

51 Magnetic Tapes Hold large volumes of data and provide high transfer rates Few GB for DAT (Digital Audio Tape) format, GB with DLT (Digital Linear Tape) format, 100 – 400 GB+ with Ultrium format, and 330 GB with Ampex helical scan format Transfer rates from few to 10s of MB/s Currently the cheapest storage medium Tapes are cheap, but cost of drives is very high Very slow access time in comparison to magnetic disks and optical disks limited to sequential access. Some formats (Accelis) provide faster seek (10s of seconds) at cost of lower capacity Used mainly for backup, for storage of infrequently used information, and as an off-line medium for transferring information from one system to another. Tape jukeboxes used for very large capacity storage (terabyte (1012 bytes) to petabye (1015 bytes)

52 Storage Access A database file is partitioned into fixed-length storage units called blocks. Blocks are units of both storage allocation and data transfer. Database system seeks to minimize the number of block transfers between the disk and memory. We can reduce the number of disk accesses by keeping as many blocks as possible in main memory. Buffer – portion of main memory available to store copies of disk blocks. Buffer manager – subsystem responsible for allocating buffer space in main memory.

53 Buffer Manager Programs call on the buffer manager when they need a block from disk. Buffer manager does the following: If the block is already in the buffer, return the address of the block in main memory If the block is not in the buffer Allocate space in the buffer for the block Replacing (throwing out) some other block, if required, to make space for the new block. Replaced block should be written back to disk only if it was modified since the most recent time that it was written to/fetched from the disk. Read the block from the disk to the buffer, and return the address of the block in main memory to requester.

54 Buffer-Replacement Policies (1/2)Most operating systems replace the block least recently used (LRU strategy) LRU is also popular in database applications Queries have well-defined access patterns (such as sequential scans), and a database system can use the information in a user’s query to predict future references LRU can be a bad strategy for certain access patterns involving repeated scans of data e.g. when computing the join of 2 relations r and s by a nested loops for each tuple tr of r do for each tuple ts of s do if the tuples tr and ts match …

55 Buffer-Replacement Policies (2/2)Most recently used (MRU) strategy – system must pin the block currently being processed. After the final tuple of that block has been processed, the block is unpinned, and it becomes the most recently used block. Buffer manager can use statistical information regarding the probability that a request will reference a particular relation For example, the data dictionary is frequently accessed. Heuristic: keep data-dictionary blocks in main memory buffer Mixed strategy with hints on replacement strategy provided by the query optimizer is preferable No single strategy is known that handles all the possible scenarios well !!!

56 Other issues Pinned block – memory block that is not allowed to be written back to disk. Toss-immediate strategy – frees the space occupied by a block as soon as the final tuple of that block has been processed Buffer managers also support forced output of blocks for the purpose of recovery (more in Chapter 17)

57 File Organization The database is stored as a collection of files. Each file is a sequence of records. A record is a sequence of fields. One approach: assume record size is fixed each file has records of one particular type only different files are used for different relations This case is easiest to implement; will consider variable length records later.

58 Fixed-Length Records Simple approach:Store record i starting from byte n  (i – 1), where n is the size of each record. Record access is simple but records may cross blocks Modification: do not allow records to cross block boundaries Deletion of record i: alternatives: move records i + 1, . . ., n to i, , n – 1 move record n to i do not move records, but link all free records on a free list

59 Free Lists Store the address of the first deleted record in the file header. Use this first record to store the address of the second deleted record, and so on Can think of these stored addresses as pointers since they “point” to the location of a record. More space efficient representation: reuse space for normal attributes of free records to store pointers. (No pointers stored in in-use records.)

60 Variable-Length RecordsVariable-length records arise in database systems in several ways: Storage of multiple record types in a file. Record types that allow variable lengths for one or more fields. Record types that allow repeating fields (used in some older data models). The slotted-page structure is commonly used in practice

61 Slotted Page Structure (1/2)Slotted page header contains: number of record entries end of free space in the block location and size of each record Records can be moved around within a page to keep them contiguous with no empty space between them; entry in the header must be updated.

62 Slotted Page Structure (2/2)Header could contain Owner, Page id, page creation time, … LMT (last modification time for LRU), LSN (log sequence number) and much more Pointers should not point directly to record — instead they should point to the entry for the record in header. Commonly, Tuple id (or object id) consists of “page id + slot number” Such indirection offers great merit !!! DBMS can reorganize the page anytime, without worrying that page id becomes “dangled”

63 Large Objects Most DBMSs restrict the size of a record to be no larger than the size of a block Remember that SQL supports types BLOB, CLOB Large objects are stored in separate, special files, not with other (short) attributes

64 Organization of Records in FilesHeap – a record can be placed anywhere in the file where there is space Sequential – store records in sequential order, based on the value of the search key of each record Hashing – a hash function computed on some attribute of each record; the result specifies in which block of the file the record should be placed Records of each relation may be stored in a separate file. In a multitable clustering file organization records of several different relations can be stored in the same file Motivation: store related records on the same block to minimize I/O

65 Sequential File Organization (1/3)Suitable for applications that require sequential processing of the entire file The records in the file are ordered by a search-key

66 Sequential File Organization (2/3)To minimize the number of block accesses in sequential file processing, we store records physically in search-key order or as close to search-key order as possible It is difficult to maintain physical sequential order !!! Very costly to move records every time inserted or deleted

67 Sequential File Organization (3/3)Deletion – use pointer chains Insertion –locate the position where the record is to be inserted if there is free space insert there if no free space, insert the record in an overflow block In either case, pointer chain must be updated Need to reorganize the file from time to time to restore sequential order

68 Multitable Clustering File Organization (1/2)Store several relations in one file using a multitable clustering file organization

69 depositor, customer, Multitable clustering

70 Multitable Clustering File Organization (2/2)Stores related records of two or more relations in each block good for queries involving depositor customer, and for queries involving one single customer and his accounts bad for queries involving only customer, select * from customer; Instead of several customer records appearing in one block, each record is located in a distinct block See the next slide.

71 Clustering File Structure With Pointer Chains

72 Oracle Example: Cluster (1/2)Create Clusters CREATE CLUSTER emp_dept (deptno NUMBER(3)) SIZE 600 TABLESPACE users STORAGE ( INITIAL 200K NEXT 300K MINEXTENTS PCTINCREASE 33); Create Clustered Tables CREATE TABLE emp (empno NUMBER(5) PRIMARY KEY, ename VARCHAR2(15) NOT NULL, deptno NUMBER(3) REFERENCES dept) CLUSTER emp_dept (deptno); CREATE TABLE dept (deptno NUMBER(3) PRIMARY KEY, ) CLUSTER emp_dept (deptno);

73 ORACLE Example : Cluster (2/2)

74 Data Dictionary StorageData dictionary (also called system catalog) stores metadata: that is, data about data Information about relations names of relations names and types of attributes of each relation names and definitions of views integrity constraints User and accounting information, including passwords Statistical and descriptive data number of tuples in each relation Physical file organization information How relation is stored (sequential/hash/…) Physical location of relation Information about indices (Chapter 12)

75 Data Dictionary Storage (Cont.)Catalog structure Relational representation on disk specialized data structures designed for efficient access, in memory A possible catalog representation: Relation_metadata = (relation_name, number_of_attributes, storage_organization, location) Attribute_metadata = (relation_name, attribute_name, domain_type, position, length) User_metadata = (user_name, encrypted_password, group) Index_metadata = (relation_name, index_name, index_type, index_attributes) View_metadata = (view_name, definition)

76 End of Chapter

77 Record RepresentationRecords with fixed length fields are easy to represent Similar to records (structs) in programming languages Extensions to represent null values E.g. a bitmap indicating which attributes are null Variable length fields can be represented by a pair (offset, length) offset: the location within the record, length: field length. All fields start at predefined location, but extra indirection required for variable length fields A-102 10 400 000 Perryridge account_number balance branch_name null-bitmap Example record structure of account record

78 Byte-String Representation of Variable-Length Records

79 Byte-String Representation of Variable-Length RecordsAttach an end-of-record () control character to the end of each record Difficulty with deletion Difficulty with growth

80 Fixed-Length RepresentationUse one or more fixed length records: reserved space pointers Reserved space – can use fixed-length records of a known maximum length; unused space in shorter records filled with a null or end-of-record symbol.

81 Pointer Method Pointer methodA variable-length record is represented by a list of fixed-length records, chained together via pointers. Can be used even if the maximum record length is not known

82 Pointer Method (Cont.) Disadvantage to pointer structure; space is wasted in all records except the first in a a chain. Solution is to allow two kinds of block in file: Anchor block – contains the first records of chain Overflow block – contains records other than those that are the first records of chairs.

83 Mapping of Objects to FilesMapping objects to files is similar to mapping tuples to files in a relational system; object data can be stored using file structures. Objects in O-O databases may lack uniformity and may be very large; such objects have to managed differently from records in a relational system. Set fields with a small number of elements may be implemented using data structures such as linked lists. Set fields with a larger number of elements may be implemented as separate relations in the database. Set fields can also be eliminated at the storage level by normalization. Similar to conversion of multivalued attributes of E-R diagrams to relations

84 Mapping of Objects to Files (Cont.)Objects are identified by an object identifier (OID); the storage system needs a mechanism to locate an object given its OID (this action is called dereferencing). logical identifiers do not directly specify an object’s physical location; must maintain an index that maps an OID to the object’s actual location. physical identifiers encode the location of the object so the object can be found directly. Physical OIDs typically have the following parts: 1. a volume or file identifier 2. a page identifier within the volume or file 3. an offset within the page

85 Management of Persistent PointersPhysical OIDs may be a unique identifier. This identifier is stored in the object also and is used to detect references via dangling pointers.

86 Management of Persistent Pointers (Cont.)Implement persistent pointers using OIDs; persistent pointers are substantially longer than are in-memory pointers Pointer swizzling cuts down on cost of locating persistent objects already in-memory. Software swizzling (swizzling on pointer deference) When a persistent pointer is first dereferenced, the pointer is swizzled (replaced by an in-memory pointer) after the object is located in memory. Subsequent dereferences of of the same pointer become cheap. The physical location of an object in memory must not change if swizzled pointers pont to it; the solution is to pin pages in memory When an object is written back to disk, any swizzled pointers it contains need to be unswizzled.

87 Hardware Swizzling With hardware swizzling, persistent pointers in objects need the same amount of space as in-memory pointers — extra storage external to the object is used to store rest of pointer information. Uses virtual memory translation mechanism to efficiently and transparently convert between persistent pointers and in-memory pointers. All persistent pointers in a page are swizzled when the page is first read in. thus programmers have to work with just one type of pointer, i.e., in-memory pointer. some of the swizzled pointers may point to virtual memory addresses that are currently not allocated any real memory (and do not contain valid data)

88 Hardware Swizzling Persistent pointer is conceptually split into two parts: a page identifier, and an offset within the page. The page identifier in a pointer is a short indirect pointer: Each page has a translation table that provides a mapping from the short page identifiers to full database page identifiers. Translation table for a page is small (at most 1024 pointers in a 4096 byte page with 4 byte pointer) Multiple pointers in page to the same page share same entry in the translation table.

89 Hardware Swizzling (Cont.)Page image before swizzling (page located on disk)

90 Hardware Swizzling (Cont.)When system loads a page into memory the persistent pointers in the page are swizzled as described below Persistent pointers in each object in the page are located using object type information For each persistent pointer (pi, oi) find its full page ID Pi If Pi does not already have a virtual memory page allocated to it, allocate a virtual memory page to Pi and read-protect the page Note: there need not be any physical space (whether in memory or on disk swap-space) allocated for the virtual memory page at this point. Space can be allocated later if (and when) Pi is accessed. In this case read-protection is not required. Accessing a memory location in the page in the will result in a segmentation violation, which is handled as described later Let vi be the virtual page allocated to Pi (either earlier or above) Replace (pi, oi) by (vi, oi) Replace each entry (pi, Pi) in the translation table, by (vi, Pi)

91 Hardware Swizzling (Cont.)When an in-memory pointer is dereferenced, if the operating system detects the page it points to has not yet been allocated storage, or is read-protected, a segmentation violation occurs. The mmap() call in Unix is used to specify a function to be invoked on segmentation violation The function does the following when it is invoked Allocate storage (swap-space) for the page containing the referenced address, if storage has not been allocated earlier. Turn off read-protection Read in the page from disk Perform pointer swizzling for each persistent pointer in the page, as described earlier

92 Hardware Swizzling (Cont.)Page image after swizzling Page with short page identifier 2395 was allocated address Observe change in pointers and translation table. Page with short page identifier 4867 has been allocated address No change in pointer and translation table.

93 Hardware Swizzling (Cont.)After swizzling, all short page identifiers point to virtual memory addresses allocated for the corresponding pages functions accessing the objects are not even aware that it has persistent pointers, and do not need to be changed in any way! can reuse existing code and libraries that use in-memory pointers After this, the pointer dereference that triggered the swizzling can continue Optimizations: If all pages are allocated the same address as in the short page identifier, no changes required in the page! No need for deswizzling — swizzled page can be saved as-is to disk A set of pages (segment) can share one translation table. Pages can still be swizzled as and when fetched (old copy of translation table is needed). A process should not access more pages than size of virtual memory — reuse of virtual memory addresses for other pages is expensive

94 Disk versus Memory Structure of ObjectsThe format in which objects are stored in memory may be different from the formal in which they are stored on disk in the database. Reasons are: software swizzling – structure of persistent and in-memory pointers are different database accessible from different machines, with different data representations Make the physical representation of objects in the database independent of the machine and the compiler. Can transparently convert from disk representation to form required on the specific machine, language, and compiler, when the object (or page) is brought into memory.

95 Large Objects Large objects : binary large objects (blobs) and character large objects (clobs) Examples include: text documents graphical data such as images and computer aided designs audio and video data Large objects may need to be stored in a contiguous sequence of bytes when brought into memory. If an object is bigger than a page, contiguous pages of the buffer pool must be allocated to store it. May be preferable to disallow direct access to data, and only allow access through a file-system-like API, to remove need for contiguous storage.

96 Modifying Large ObjectsIf the application requires insert/delete of bytes from specified regions of an object: B+-tree file organization (described later in Chapter 12) can be modified to represent large objects Each leaf page of the tree stores between half and 1 page worth of data from the object Special-purpose application programs outside the database are used to manipulate large objects: Text data treated as a byte string manipulated by editors and formatters. Graphical data and audio/video data is typically created and displayed by separate application checkout/checkin method for concurrency control and creation of versions