File Syst ems Last t ime we t alked about disk int ernals 11: - PDF document

File Syst ems � Last t ime we t alked about disk int ernals 11: File Syst em Basics � Despit e complex int ernals, disks export a simple array of sect ors Last Modif ied: � How do we go f rom t hat t o a f ile syst em? 6/ 15/ 2004 12:12:09 PM � What do we exact ly do we expect f rom a f ile syst em? -1 -2 File Syst em Basics Files � FS are probably t he OS abst ract ion t hat � A f ile is a collect ion of dat a wit h syst em maint ained pr oper t ies like average user is most f amiliar wit h � Owner, size, name, last read/ writ e t ime, et c. � Files � Files of t en have “t ypes” which allow user s and � Dir ect or ies applicat ions t o r ecognize t heir int ended use � Access cont r ols (owner s, gr oups, per missions) � Some f ile t ypes ar e under st ood by t he f ile syst em (mount point , symbolic link, dir ect or y) � Some f ile t ypes ar e under st ood by applicat ions and user s (.t xt , .j pg, .ht ml, .doc, … ) � Could t he syst em underst and t hese t ypes and cust omize it s handling? -3 -4 Basic File Operat ions Direct ories Windows � Dir ect or ies pr ovide a way f or user s t o or ganize UNI X t heir f iles *and* a convenient way f or user s t o � Creat eFile (name, CREATE) � cr eat e (name) ident if y and shar e dat a � Creat eFile (name, OPEN) � open (name, mode) � Logically dir ect or ies st or e inf or mat ion like f ile � ReadFile (handle) � r ead (f d) name, size, modif icat ion t ime et c (Not always kept � Wr it eFile (handle) in t he dir ect or y t hough..) � writ e(f d) � FlushFileBuf f er s(handle) � Most f ile syst ems suppor t hier ar chical dir ect or ies � sync(f d) � Set FilePoint er (handle) (/ usr/ local/ bin or C:\ WI NNT) � seek(f d, pos) � CloseHandle(handl ) � P eople like t o organize inf ormat ion hierarchically � close(f d) � Delet eFile(name) � Recall: OS of t en r ecor ds a cur r ent wor king dir ect or y f or each pr ocess � unlink (name) � CopyFile (name) � Can t heref ore ref er t o f iles by absolut e and relat ive � r ename (old, new) � MoveFile(name) names -5 -6 1

Direct ories are special f iles Mult iple parent direct ories? � Dir ect or ies ar e f iles cont aining inf or mat ion t o be � One nat ur al quest ion is “can a f ile be in mor e t han int er pr et ed by t he f ile syst em it self one dir ect or y”? � List of f iles and ot her direct ories cont ained in t his � Sof t links direct ory � Special f ile int erpret ed by t he FS (like direct ories in � Some at t ribut es of each child including where t o f ind it !! t hat sense) � How should t he list of childr en be or ganized? � Tell FS t o look at a dif f erent pat hname f or t his f ile � Flat f ile? � I f f ile delet ed or moved, sof t link will point t o wrong place � B- tree? � Har d links � Many syst ems have no par t icular or der , but t his is � Along wit h ot her f ile inf o maint ain ref erence count ext r emely bad f or lar ge dir ect or ies! � Delet e f ile = decrement ref erence count � Only reclaim st orage when ref erence count does t o 0 -7 -8 P at h Name Translat ion File Buf f er Cache � Cache Dat a Read � To f ind f ile “/ f oo/ bar/ baz” � Exploit t emporal localit y of access by caching pat hname � Find t he special root direct ory f ile (how does FS know t ranslat ion inf ormat ion where t hat is?) � Exploit t emporal localit y of access by leaving recent ly � I n special root direct ory f ile, look f or ent ry f oo and t hat accesses chunks of a f ile in memory in hopes t hat t hey ent ry will t ell you where f oo is will be accessed again (let app give hint if not ?) � Read special direct ory f ile f oo and look f or ent ry bar t o � Exploit spat ial localit y of access by bringing in large t ell you where bar is chunks of a f ile at once � Find special direct ory f ile bar and look f or ent ry baz t o � Dat a wr it t en is also cached t ell you where baz is � For correct ness should be writ e-t hrough t o disk � Finally f ind baz � Normally is writ e- behind � FS can cache common pr ef ixes f or ef f iciency • FS per iodically walks t he buf f er cache and “f lushes” t hings older t han 30 seconds t o disk • Unr eliable! � Usually LRU r eplacement -9 -10 File Buf f er Cache P rot ect ion Syst em � Most FS implement a pr ot ect ion scheme t o � Typically cache is syst em wide (shared by cont rol: all processes) � Who can access a f ile � Shar ed libr ar ies and execut ables and ot her � How t hey can access it (e.g. read/ writ e/ exec/ ..) commonly accessed f iles likely t o be in memor y � Any pr ot ect ion syst em dict at es whet her a given already act ion per f or med by a given subj ect on a given obj ect should be allowed. I n t his case: � Compet es wit h virt ual memory syst em f or physical memory � Obj ect s = f iles � Processes have less memory available t o t hem � P rinciples = users t o st or e code and dat a (addr ess space) � Act ions = operat ions � We’ll t alk mor e about pr ot ect ion syst ems lat er in � Some syst ems have int egr at ed VM/ FS caches t he semest er -11 -12 2

Exercise f or t he Reader ☺ File Syst ems � We t alked a bit about disk int ernals � I f you wer e going t o build your own f ile syst em on t op of a f ixed sized f ile what � Despit e complex int ernals, disks export a would you do? simple array of sect ors � What ot her inf or mat ion would you need t o st or e t her e besides f ile dat a and dir ect or y � How do we go f rom t hat t o a f ile syst em? dat a? � How would you or ganize t hings? -13 -14 Some quest ions File Layout � Would you keep each f ile t oget her � Opt ion 1: All blocks in a f ile must be sequent ially? allocat ed cont iguously � Only need t o list st art and lengt h in direct ory � I f you did, what would you do if a f ile gr ew or � Causes f r agment at ion of f r ee space shrunk? � Also causes copying as f iles gr ow � I f not , how would you keep t r ack of t he � Opt ion 2: Allow f iles t o be broken int o mult iple pieces? pieces � Fixed sized pieces (blocks) or var iable sized pieces (ext ent s)? � I f we ar e going t o allow f iles t o be br oken int o mult iple pieces how will we keep t r ack of t hem ? -15 -16 Blocks or Ext ent s? Finding all P art s of a File � I f f ixed sized block t hen st ore j ust � Opt ion 2A: List all blocks in t he dir ect or y � Direct ories will get pret t y big and also must change t he st ar t ing locat ion f or each one direct ory everyt ime a f ile grows or shrinks � I f variable sized ext ent need t o st ore � Opt ion 2B: Linked st r uct ur e st art ing locat ion and lengt h � Direct ory point s t o f irst piece (block or ext ent ), f irst one point s t o second one � But maybe you can have f ewer ext ent s? � File can expand and cont ract wit hout copying � Blocks = less ext ernal f ragment at ion � Good f or sequent ial access, t errible f or ot her kinds � Ext ent s = less int ernal f ragment at ion � Opt ion 2C: I ndexed st r uct ur e � Direct ory point s t o index block which cont ains point ers t o dat a blocks � Good f or random access as well as sequent ial access -17 -18 3