// Berkeley Open Infrastructure for Network Computing // http://boinc.berkeley.edu // Copyright (C) 2005 University of California // // This is free software; you can redistribute it and/or // modify it under the terms of the GNU Lesser General Public // License as published by the Free Software Foundation; // either version 2.1 of the License, or (at your option) any later version. // // This software is distributed in the hope that it will be useful, // but WITHOUT ANY WARRANTY; without even the implied warranty of // MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. // See the GNU Lesser General Public License for more details. // // To view the GNU Lesser General Public License visit // http://www.gnu.org/copyleft/lesser.html // or write to the Free Software Foundation, Inc., // 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA #ifndef _BOINC_DB_ #define _BOINC_DB_ // Structures corresponding to database records. // Some of these types have counterparts in client/types.h, // but don't be deceived - client and server have different variants. // The parse and write functions are for use in scheduler RPC. // They don't necessarily serialize the entire records. #include #include #include "db_base.h" extern DB_CONN boinc_db; // Sizes of text buffers in memory, corresponding to database BLOBs. // Large is for fields with user-supplied text, and preferences //#define MEDIUM_BLOB_SIZE 4096 #define LARGE_BLOB_SIZE 65536 // Dummy name for file xfers #define FILE_MOVER "move_file" // A compilation target, i.e. a architecture/OS combination. // The core client will be given only applications with the same platform // struct PLATFORM { int id; int create_time; char name[256]; // i.e. "sparc-sun-solaris" char user_friendly_name[256]; // i.e. "SPARC Solaris 2.8" int deprecated; void clear(); }; // A version of the core client // struct CORE_VERSION { int id; int create_time; int version_num; int platformid; char xml_doc[LARGE_BLOB_SIZE]; // a for the download file char message[256]; // if we get a request from this version, // send this message bool deprecated; // if we get a request from this version, // don't send it any work. void clear(); }; // An application. // struct APP { int id; int create_time; char name[256]; // application name, preferably short int min_version; // don't use app versions before this bool deprecated; char user_friendly_name[256]; bool homogeneous_redundancy; int write(FILE*); void clear(); }; // A version of an application. // struct APP_VERSION { int id; int create_time; int appid; int version_num; int platformid; char xml_doc[LARGE_BLOB_SIZE]; // describes app files. format: // ... // ... // // ... // x // // ... // [] // [] // // // // the following let you handle backwards-incompatible changes to // the core client / app interface // int min_core_version; // min core version this app will run with int max_core_version; // if <>0, max core version this will run with bool deprecated; int write(FILE*, APP&); void clear(); }; struct USER { int id; int create_time; char email_addr[256]; char name[256]; char authenticator[256]; char country[256]; char postal_code[256]; double total_credit; double expavg_credit; // credit per second, recent average double expavg_time; // when the above was computed char global_prefs[LARGE_BLOB_SIZE]; // global preferences, within tag char project_prefs[LARGE_BLOB_SIZE]; // project preferences; format: // // X // // ... // // // x // // ... // // // ... // int teamid; // team ID if any char venue[256]; // home/work/school (default) char url[256]; // user's web page if any bool send_email; bool show_hosts; int posts; // number of messages posted (redundant) // The following are specific to SETI@home; // they record info about the user's involvement in a prior project int seti_id; // ID in old DB int seti_nresults; // number of WUs completed int seti_last_result_time; // time of last result (UNIX) double seti_total_cpu; // number of CPU seconds char signature[256]; bool has_profile; char cross_project_id[256]; void clear(); }; #define TEAM_TYPE_CLUB 1 #define TEAM_TYPE_COMPANY 2 #define TEAM_TYPE_PRIMARY 3 #define TEAM_TYPE_SECONDARY 4 #define TEAM_TYPE_JUNIOR_COLLEGE 5 #define TEAM_TYPE_UNIVERSITY 6 #define TEAM_TYPE_GOVERNMENT 7 // invariants of teams: // a team has > 0 members struct TEAM { int id; int create_time; int userid; // User ID of team founder char name[256]; char name_lc[256]; // Team name in lowercase (used for searching) char url[256]; int type; // Team type (see above) char name_html[256]; char description[LARGE_BLOB_SIZE]; int nusers; // UNDEFINED BY DEFAULT char country[256]; double total_credit; double expavg_credit; double expavg_time; // The following is specific to SETI@home int seti_id; // ID in old DB void clear(); }; struct HOST { int id; int create_time; int userid; // ID of user running this host int rpc_seqno; // last seqno received from client int rpc_time; // time of last scheduler RPC double total_credit; double expavg_credit; // credit per second, recent average double expavg_time; // last time the above was updated // all remaining items are assigned by the client int timezone; // hours difference from GMT char domain_name[256]; char serialnum[256]; char last_ip_addr[256]; // internal IP address as of last RPC int nsame_ip_addr; // # of RPCs with same IP address double on_frac; // see client/time_stats.h double connected_frac; double active_frac; int p_ncpus; // Number of CPUs on host char p_vendor[256]; // Vendor name of CPU char p_model[256]; // Model of CPU double p_fpops; // measured floating point ops/sec of CPU double p_iops; // measured integer ops/sec of CPU double p_membw; // measured memory bandwidth (bytes/sec) of CPU // The above are per CPU, not total double p_calculated; // when the above were calculated char os_name[256]; // Name of operating system char os_version[256]; // Version of operating system double m_nbytes; // Size of memory in bytes double m_cache; // Size of CPU cache in bytes (L1 or L2?) double m_swap; // Size of swap space in bytes double d_total; // Total disk space on host // - may include all volumes, // even if BOINC can use only one of them // - may include network (shared) storage double d_free; // Of the total disk space, how much is free double d_boinc_used_total; // amount being used for all projects double d_boinc_used_project; // amount being used for this project double d_boinc_max; // max amount that BOINC is allowed to use // This reflects both user preferences // and the fact that BOINC can use only 1 volume double n_bwup; // Average upload bandwidth, bytes/sec double n_bwdown; // Average download bandwidth, bytes/sec // The above are derived from actual // file upload/download times, and may reflect // factors other than network bandwidth // The following is derived (by server) from other fields double credit_per_cpu_sec; char venue[256]; // home/work/school int nresults_today; // results sent since midnight double avg_turnaround; // recent average result turnaround time char host_cpid[256]; // host cross-project ID char external_ip_addr[256]; // IP address seen by scheduler int parse(FILE*); int parse_time_stats(FILE*); int parse_net_stats(FILE*); void fix_nans(); void clear(); }; // values for file_delete state #define FILE_DELETE_INIT 0 #define FILE_DELETE_READY 1 // set to this value only when we believe all files are uploaded #define FILE_DELETE_DONE 2 // means the file uploader ATTEMPTED to delete files. // May have failed. TODO: retry delete later // values for assimilate_state #define ASSIMILATE_INIT 0 #define ASSIMILATE_READY 1 #define ASSIMILATE_DONE 2 // NOTE: there is no overall state for a WU (like done/not done) // There's just a bunch of independent substates // (file delete, assimilate, and states of results, error flags) // bit fields of error_mask #define WU_ERROR_COULDNT_SEND_RESULT 1 #define WU_ERROR_TOO_MANY_ERROR_RESULTS 2 #define WU_ERROR_TOO_MANY_SUCCESS_RESULTS 4 #define WU_ERROR_TOO_MANY_TOTAL_RESULTS 8 #define WU_ERROR_CANCELLED 16 struct WORKUNIT { int id; int create_time; int appid; // associated app char name[256]; char xml_doc[LARGE_BLOB_SIZE]; int batch; double rsc_fpops_est; // estimated # of FP operations // used to estimate how long a result will take on a host double rsc_fpops_bound; // upper bound on # of FP ops // used to calculate an upper bound on the CPU time for a result // before it is aborted. double rsc_memory_bound; // upper bound on RAM working set (bytes) // currently used only by scheduler to screen hosts // At some point, could use as runtime limit double rsc_disk_bound; // upper bound on amount of disk needed (bytes) // (including input, output and temp files, but NOT the app) // used for 2 purposes: // 1) for scheduling (don't send this WU to a host w/ insuff. disk) // 2) abort task if it uses more than this disk bool need_validate; // this WU has at least 1 result in // validate state = NEED_CHECK int canonical_resultid; // ID of canonical result, or zero double canonical_credit; // credit that all correct results get int transition_time; // when should transition_handler // next check this WU? // MAXINT if no need to check int delay_bound; // determines result deadline, // timeout check time int error_mask; // bitmask of errors (see above) int file_delete_state; int assimilate_state; int hr_class; // homogeneous redundancy class // used to send redundant copies only to "similar" hosts // (in terms of numerics, performance, or both) double opaque; // project-specific; usually external ID int min_quorum; // minimum quorum size int target_nresults; // try to get this many successful results // may be > min_quorum to get consensus // quicker or reflect loss rate int max_error_results; // WU error if < #error results int max_total_results; // WU error if < #total results // (need this in case results are never returned) int max_success_results; // WU error if < #success results // without consensus (i.e. WU is nondeterministic) char result_template_file[64]; int priority; char mod_time[16]; // the following not used in the DB char app_name[256]; void clear(); }; // WARNING: be Very careful about changing any values, // especially for a project already running - // the database will become inconsistent #define RESULT_SERVER_STATE_INACTIVE 1 #define RESULT_SERVER_STATE_UNSENT 2 #define RESULT_SERVER_STATE_UNSENT_SEQ 3 // unsent, part of a work sequence #define RESULT_SERVER_STATE_IN_PROGRESS 4 #define RESULT_SERVER_STATE_OVER 5 // we received a reply, timed out, or decided not to send. // Note: we could get a reply even after timing out. #define RESULT_OUTCOME_INIT 0 #define RESULT_OUTCOME_SUCCESS 1 #define RESULT_OUTCOME_COULDNT_SEND 2 #define RESULT_OUTCOME_CLIENT_ERROR 3 // an error happened on the client #define RESULT_OUTCOME_NO_REPLY 4 #define RESULT_OUTCOME_DIDNT_NEED 5 // we created the result but didn't need to send it because we already // got a quorum #define RESULT_OUTCOME_VALIDATE_ERROR 6 // The outcome was initially SUCCESS, but the validator // had a permanent error reading a result file, // or the result file had a syntax error #define VALIDATE_STATE_INIT 0 #define VALIDATE_STATE_VALID 1 #define VALIDATE_STATE_INVALID 2 #define VALIDATE_STATE_NO_CHECK 3 // WU had error, so we'll never get around to validating its results // This lets us avoid showing the claimed credit as "pending" #define VALIDATE_STATE_INCONCLUSIVE 4 // the validator looked this result (as part of a check_set() call) // but didn't find a canonical result. // This needs to be distinct from INIT for the transitioner to decide // whether to trigger the validator #define VALIDATE_STATE_TOO_LATE 5 // The result arrived after the canonical result's files were deleted, // so we can't determine if it's valid struct RESULT { int id; int create_time; int workunitid; int server_state; // see above int outcome; // see above; defined if server state OVER int client_state; // phase that client contacted us in. // if UPLOADED then outcome is success. // error details are in stderr_out. // The values for this field are defined // in lib/result_state.h int hostid; // host processing this result int userid; // user processing this result int report_deadline; // deadline for receiving result int sent_time; // when result was sent to host int received_time; // when result was received from host char name[256]; double cpu_time; // CPU time used to complete result char xml_doc_in[LARGE_BLOB_SIZE]; // descriptions of output files char xml_doc_out[LARGE_BLOB_SIZE]; // MD5s of output files char stderr_out[LARGE_BLOB_SIZE]; // stderr output, if any int batch; int file_delete_state; // see above; values for file_delete_state int validate_state; double claimed_credit; // CPU time times host credit/sec double granted_credit; // == canonical credit of WU double opaque; // project-specific; usually external ID int random; // determines send order int app_version_num; // version# of app (not core client) int appid; // copy of WU's appid int exit_status; // application exit status, if any int teamid; int priority; char mod_time[16]; // the following not used in the DB char wu_name[256]; int parse_from_client(FILE*); void clear(); }; struct MSG_FROM_HOST { int id; int create_time; int hostid; char variety[256]; // project-defined; what kind of msg bool handled; // message handler has processed this char xml[LARGE_BLOB_SIZE]; void clear(); }; struct MSG_TO_HOST { int id; int create_time; int hostid; char variety[256]; // project-defined; what kind of msg bool handled; // scheduler has sent this char xml[LARGE_BLOB_SIZE]; // text to include in sched reply void clear(); }; struct TRANSITIONER_ITEM { int id; char name[256]; int appid; int min_quorum; bool need_validate; int canonical_resultid; int transition_time; int delay_bound; int error_mask; int max_error_results; int max_total_results; int file_delete_state; int assimilate_state; int target_nresults; char result_template_file[64]; int priority; int res_id; char res_name[256]; int res_report_deadline; int res_server_state; int res_outcome; int res_validate_state; int res_file_delete_state; int res_sent_time; int res_hostid; void clear(); void parse(MYSQL_ROW&); }; struct VALIDATOR_ITEM { WORKUNIT wu; RESULT res; void clear(); void parse(MYSQL_ROW&); }; class DB_PLATFORM : public DB_BASE, public PLATFORM { public: DB_PLATFORM(DB_CONN* p=0); int get_id(); void db_print(char*); void db_parse(MYSQL_ROW &row); }; class DB_CORE_VERSION : public DB_BASE, public CORE_VERSION { public: DB_CORE_VERSION(DB_CONN* p=0); int get_id(); void db_print(char*); void db_parse(MYSQL_ROW &row); }; class DB_APP : public DB_BASE, public APP { public: DB_APP(DB_CONN* p=0); int get_id(); void db_print(char*); void db_parse(MYSQL_ROW &row); }; class DB_APP_VERSION : public DB_BASE, public APP_VERSION { public: DB_APP_VERSION(DB_CONN* p=0); int get_id(); void db_print(char*); void db_parse(MYSQL_ROW &row); }; class DB_USER : public DB_BASE, public USER { public: DB_USER(DB_CONN* p=0); int get_id(); void db_print(char*); void db_parse(MYSQL_ROW &row); void operator=(USER& r) {USER::operator=(r);} }; class DB_TEAM : public DB_BASE, public TEAM { public: DB_TEAM(DB_CONN* p=0); int get_id(); void db_print(char*); void db_parse(MYSQL_ROW &row); }; class DB_HOST : public DB_BASE, public HOST { public: DB_HOST(DB_CONN* p=0); int get_id(); void db_print(char*); void db_parse(MYSQL_ROW &row); void operator=(HOST& r) {HOST::operator=(r);} }; class DB_RESULT : public DB_BASE, public RESULT { public: DB_RESULT(DB_CONN* p=0); int get_id(); int update_subset(); void db_print(char*); void db_print_values(char*); void db_parse(MYSQL_ROW &row); void operator=(RESULT& r) {RESULT::operator=(r);} }; class DB_WORKUNIT : public DB_BASE, public WORKUNIT { public: DB_WORKUNIT(DB_CONN* p=0); int get_id(); void db_print(char*); void db_parse(MYSQL_ROW &row); void operator=(WORKUNIT& w) {WORKUNIT::operator=(w);} }; class DB_MSG_FROM_HOST : public DB_BASE, public MSG_FROM_HOST { public: DB_MSG_FROM_HOST(DB_CONN* p=0); int get_id(); void db_print(char*); void db_parse(MYSQL_ROW &row); }; class DB_MSG_TO_HOST : public DB_BASE, public MSG_TO_HOST { public: DB_MSG_TO_HOST(DB_CONN* p=0); int get_id(); void db_print(char*); void db_parse(MYSQL_ROW &row); }; // The transitioner uses this to get (WU, result) pairs efficiently. // Each call to enumerate() returns a list of the pairs for a single WU // class DB_TRANSITIONER_ITEM_SET : public DB_BASE_SPECIAL { public: DB_TRANSITIONER_ITEM_SET(DB_CONN* p=0); TRANSITIONER_ITEM last_item; int nitems_this_query; int enumerate( int transition_time, int nresult_limit, std::vector& items ); int update_result(TRANSITIONER_ITEM&); int update_workunit(TRANSITIONER_ITEM&); }; // The validator uses this to get (WU, result) pairs efficiently. // Each call to enumerate() returns a list of the pairs for a single WU // class DB_VALIDATOR_ITEM_SET : public DB_BASE_SPECIAL { public: DB_VALIDATOR_ITEM_SET(DB_CONN* p=0); VALIDATOR_ITEM last_item; int nitems_this_query; int enumerate( int appid, int nresult_limit, std::vector& items ); int update_result(RESULT&); int update_workunit(WORKUNIT&); }; // used by the feeder and scheduler for outgoing work // struct WORK_ITEM { int res_id; WORKUNIT wu; void parse(MYSQL_ROW& row); }; class DB_WORK_ITEM : public WORK_ITEM, public DB_BASE_SPECIAL { public: DB_WORK_ITEM(DB_CONN* p=0); // CURSOR cursor; int enumerate(int limit, char* order_clause); // used by feeder int read_result(); // used by scheduler to read result server state int update(); // used by scheduler to update WU transition time // and various result fields }; // Used by the scheduler to handle results reported by clients // The read and the update of these results are combined // into single SQL queries. struct SCHED_RESULT_ITEM { char queried_name[256]; // name as reported by client int id; char name[256]; int workunitid; int server_state; int client_state; int validate_state; int outcome; int hostid; int userid; int teamid; int received_time; double cpu_time; double claimed_credit; char xml_doc_out[LARGE_BLOB_SIZE]; char stderr_out[LARGE_BLOB_SIZE]; int app_version_num; int exit_status; void clear(); void parse(MYSQL_ROW& row); }; class DB_SCHED_RESULT_ITEM_SET : public DB_BASE_SPECIAL { public: DB_SCHED_RESULT_ITEM_SET(DB_CONN* p=0); std::vector results; int add_result(char* result_name); int enumerate(); // using a single SQL query, look up all the reported results, // (based on queried_name) // and fill in the rest of the entries in the results vector int lookup_result(char* result_name, SCHED_RESULT_ITEM** result); int update_result(SCHED_RESULT_ITEM& result); int update_workunits(); }; #endif