tables.mysql 26.5 KB
Newer Older
1 2
/*
    This is MySQL version of EnsEMBL Hive database schema file.
Leo Gordon's avatar
Leo Gordon committed
3

4 5
    It has been annotated with @-tags.
    The following command is used to create HTML documentation:
6
        perl $ENSEMBL_CVS_ROOT_DIR/ensembl-production/scripts/sql2html.pl -i $ENSEMBL_CVS_ROOT_DIR/ensembl-hive/sql/tables.mysql \
7 8
             -o $ENSEMBL_CVS_ROOT_DIR/ensembl-hive/docs/hive_schema.html -d Hive -sort_headers 0 -sort_tables 0

9 10
    Adding the following line into the header of the previous output will make it look prettier (valid in rel.72):
        <link rel="stylesheet" type="text/css" media="all" href="http://static.ensembl.org/minified/eb8658698ad4d45258b954c2d3f35bad.css" />
11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39
*/


/**
@header Pipeline structure
@colour #C70C09
*/

/**
@table  analysis_base

@colour #C70C09

@desc   Each Analysis is a node of the pipeline diagram.
        It acts both as a "class" to which Jobs belong (and inherit from it certain properties)
        and as a "container" for them (Jobs of an Analysis can be blocking all Jobs of another Analysis).

@column analysis_id             a unique ID that is also a foreign key to most of the other tables
@column logic_name              the name of the Analysis object
@column module                  the Perl module name that runs this Analysis
@column parameters              a stingified hash of parameters common to all jobs of the Analysis
@column resource_class_id       link to the resource_class table
@column failed_job_tolerance    % of tolerated failed Jobs
@column max_retry_count         how many times a job of this Analysis will be retried (unless there is no point)
@column can_be_empty            if TRUE, this Analysis will not be blocking if/while it doesn't have any jobs
@column priority                an Analysis with higher priority will be more likely chosen on Worker's specialization
@column meadow_type             if defined, forces this Analysis to be run only on the given Meadow
@column analysis_capacity       if defined, limits the number of Workers of this particular Analysis that are allowed to run in parallel
*/
Leo Gordon's avatar
Leo Gordon committed
40

41
CREATE TABLE analysis_base (
Leo Gordon's avatar
Leo Gordon committed
42
    analysis_id             INTEGER     NOT NULL PRIMARY KEY AUTO_INCREMENT,
43
    logic_name              VARCHAR(255) NOT NULL,
44 45
    module                  VARCHAR(255),
    parameters              TEXT,
Leo Gordon's avatar
Leo Gordon committed
46 47 48 49 50
    resource_class_id       INTEGER     NOT NULL,
    failed_job_tolerance    INTEGER     NOT NULL DEFAULT 0,
    max_retry_count         INTEGER     NOT NULL DEFAULT 3,
    can_be_empty            SMALLINT    NOT NULL DEFAULT 0,
    priority                SMALLINT    NOT NULL DEFAULT 0,
51
    meadow_type             VARCHAR(255)         DEFAULT NULL,
Leo Gordon's avatar
Leo Gordon committed
52
    analysis_capacity       INTEGER              DEFAULT NULL,
Leo Gordon's avatar
Leo Gordon committed
53

Leo Gordon's avatar
Leo Gordon committed
54
    UNIQUE  KEY logic_name_idx (logic_name)
Leo Gordon's avatar
Leo Gordon committed
55
) COLLATE=latin1_swedish_ci ENGINE=InnoDB;
Leo Gordon's avatar
Leo Gordon committed
56 57


58 59
/**
@table  analysis_stats
Jessica Severin's avatar
Jessica Severin committed
60

61
@colour #C70C09
62

63 64 65 66
@desc   Parallel table to analysis_base which provides high level statistics on the
        state of an analysis and it's jobs.  Used to provide a fast overview, and to
        provide final approval of 'DONE' which is used by the blocking rules to determine
        when to unblock other analyses.  Also provides
67

68 69 70 71
@column analysis_id             foreign-keyed to the corresponding analysis_base entry
@column batch_size              how many jobs are claimed in one claiming operation before Worker starts executing them
@column hive_capacity           a reciprocal limiter on the number of Workers running at the same time (dependent on Workers of other Analyses)
@column status                  cached state of the Analysis
Leo Gordon's avatar
Leo Gordon committed
72

73 74 75 76 77
@column total_job_count         total number of Jobs of this Analysis
@column semaphored_job_count    number of Jobs of this Analysis that are in SEMAPHORED state
@column ready_job_count         number of Jobs of this Analysis that are in READY state
@column done_job_count          number of Jobs of this Analysis that are in DONE state
@column failed_job_count        number of Jobs of this Analysis that are in FAILED state
Jessica Severin's avatar
Jessica Severin committed
78

79 80
@column num_running_workers     number of running Workers of this Analysis
@column num_required_workers    extra number of Workers of this Analysis needed to execute all READY jobs
Jessica Severin's avatar
Jessica Severin committed
81

82 83 84
@column behaviour               whether hive_capacity is set or is dynamically calculated based on timers
@column input_capacity          used to compute hive_capacity in DYNAMIC mode
@column output_capacity         used to compute hive_capacity in DYNAMIC mode
Jessica Severin's avatar
Jessica Severin committed
85

86 87 88 89
@column avg_msec_per_job        weighted average used to compute DYNAMIC hive_capacity
@column avg_input_msec_per_job  weighted average used to compute DYNAMIC hive_capacity
@column avg_run_msec_per_job    weighted average used to compute DYNAMIC hive_capacity
@column avg_output_msec_per_job weighted average used to compute DYNAMIC hive_capacity
Jessica Severin's avatar
Jessica Severin committed
90

91 92 93
@column last_update             when this entry was last updated
@column sync_lock               a binary lock flag to prevent simultaneous updates
*/
Leo Gordon's avatar
Leo Gordon committed
94

95
CREATE TABLE analysis_stats (
Leo Gordon's avatar
Leo Gordon committed
96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121
    analysis_id             INTEGER     NOT NULL,
    batch_size              INTEGER     NOT NULL DEFAULT 1,
    hive_capacity           INTEGER              DEFAULT NULL,
    status                  ENUM('BLOCKED', 'LOADING', 'SYNCHING', 'EMPTY', 'READY', 'WORKING', 'ALL_CLAIMED', 'DONE', 'FAILED') DEFAULT 'EMPTY' NOT NULL,

    total_job_count         INTEGER     NOT NULL DEFAULT 0,
    semaphored_job_count    INTEGER     NOT NULL DEFAULT 0,
    ready_job_count         INTEGER     NOT NULL DEFAULT 0,
    done_job_count          INTEGER     NOT NULL DEFAULT 0,
    failed_job_count        INTEGER     NOT NULL DEFAULT 0,
    num_running_workers     INTEGER     NOT NULL DEFAULT 0,
    num_required_workers    INTEGER     NOT NULL DEFAULT 0,

    behaviour               ENUM('STATIC', 'DYNAMIC') NOT NULL DEFAULT 'STATIC',
    input_capacity          INTEGER     NOT NULL DEFAULT 4,
    output_capacity         INTEGER     NOT NULL DEFAULT 4,

    avg_msec_per_job        INTEGER              DEFAULT NULL,
    avg_input_msec_per_job  INTEGER              DEFAULT NULL,
    avg_run_msec_per_job    INTEGER              DEFAULT NULL,
    avg_output_msec_per_job INTEGER              DEFAULT NULL,

    last_update             TIMESTAMP                    NULL,  -- mysql's special for "TIMESTAMP DEFAULT NULL"
    sync_lock               SMALLINT    NOT NULL DEFAULT 0,

    PRIMARY KEY (analysis_id)
Leo Gordon's avatar
Leo Gordon committed
122
) COLLATE=latin1_swedish_ci ENGINE=InnoDB;
Jessica Severin's avatar
Jessica Severin committed
123 124


125 126 127 128 129 130
/**
@table  dataflow_rule

@colour #C70C09

@desc Extension of simple_rule design except that goal(to) is now in extended URL format e.g.
Leo Gordon's avatar
Leo Gordon committed
131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148
        mysql://ensadmin:<pass>@ecs2:3361/compara_hive_test?analysis.logic_name='blast_NCBI34'
        (full network address of an analysis).
        The only requirement is that there are rows in the job, analysis, dataflow_rule,
        and worker tables so that the following join works on the same database 
            WHERE analysis.analysis_id = dataflow_rule.from_analysis_id 
            AND   analysis.analysis_id = job.analysis_id
            AND   analysis.analysis_id = worker.analysis_id
        These are the rules used to create entries in the job table where the
        input_id (control data) is passed from one analysis to the next to define work.
        The analysis table will be extended so that it can specify different read and write
        databases, with the default being the database the analysis is on

@column dataflow_rule_id        internal ID
@column from_analysis_id        foreign key to analysis table analysis_id
@column branch_code             branch_code of the fan
@column funnel_dataflow_rule_id dataflow_rule_id of the semaphored funnel (is NULL by default, which means dataflow is not semaphored)
@column to_analysis_url         foreign key to net distributed analysis logic_name reference
@column input_id_template       a template for generating a new input_id (not necessarily a hashref) in this dataflow; if undefined is kept original
149
*/
Jessica Severin's avatar
Jessica Severin committed
150

151
CREATE TABLE dataflow_rule (
Leo Gordon's avatar
Leo Gordon committed
152 153 154 155 156 157
    dataflow_rule_id        INTEGER     NOT NULL PRIMARY KEY AUTO_INCREMENT,
    from_analysis_id        INTEGER     NOT NULL,
    branch_code             INTEGER     NOT NULL DEFAULT 1,
    funnel_dataflow_rule_id INTEGER              DEFAULT NULL,
    to_analysis_url         VARCHAR(255) NOT NULL DEFAULT '',
    input_id_template       TEXT                 DEFAULT NULL,
Leo Gordon's avatar
Leo Gordon committed
158

Leo Gordon's avatar
Leo Gordon committed
159
    UNIQUE  KEY (from_analysis_id, branch_code, funnel_dataflow_rule_id, to_analysis_url, input_id_template(512))
Leo Gordon's avatar
Leo Gordon committed
160
) COLLATE=latin1_swedish_ci ENGINE=InnoDB;
Jessica Severin's avatar
Jessica Severin committed
161 162


163 164
/**
@table  analysis_ctrl_rule
165

166 167 168 169 170 171 172 173 174 175 176 177 178 179
@colour #C70C09

@desc   These rules define a higher level of control.
        These rules are used to turn whole anlysis nodes on/off (READY/BLOCKED).
        If any of the condition_analyses are not 'DONE' the ctrled_analysis is set to BLOCKED.
        When all conditions become 'DONE' then ctrled_analysis is set to READY
        The workers switch the analysis.status to 'WORKING' and 'DONE'.
        But any moment if a condition goes false, the analysis is reset to BLOCKED.

@column condition_analysis_url foreign key to net distributed analysis reference
@column ctrled_analysis_id     foreign key to analysis table analysis_id
*/

CREATE TABLE analysis_ctrl_rule (
Leo Gordon's avatar
Leo Gordon committed
180 181
    condition_analysis_url  VARCHAR(255) NOT NULL DEFAULT '',
    ctrled_analysis_id      INTEGER     NOT NULL,
182 183

    UNIQUE  KEY (condition_analysis_url, ctrled_analysis_id)
Leo Gordon's avatar
Leo Gordon committed
184
) COLLATE=latin1_swedish_ci ENGINE=InnoDB;
185 186


187 188 189 190 191 192 193 194 195 196 197 198 199 200 201
/**
@header Resources
@colour #FF7504
*/

/**
@table  resource_class

@colour #FF7504

@desc   Maps between resource_class numeric IDs and unique names.

@column resource_class_id   unique ID of the ResourceClass
@column name                unique name of the ResourceClass
*/
202

203
CREATE TABLE resource_class (
204 205
    resource_class_id       INTEGER      NOT NULL PRIMARY KEY AUTO_INCREMENT,
    name                    VARCHAR(255) NOT NULL,
Leo Gordon's avatar
Leo Gordon committed
206

207
    UNIQUE  KEY(name)
Leo Gordon's avatar
Leo Gordon committed
208
) COLLATE=latin1_swedish_ci ENGINE=InnoDB;
Leo Gordon's avatar
Leo Gordon committed
209 210


211 212 213 214 215 216 217 218 219
/**
@table  resource_description

@colour #FF7504

@desc   Maps (ResourceClass, MeadowType) pair to Meadow-specific resource lines.

@column resource_class_id   foreign-keyed to the ResourceClass entry
@column meadow_type         if the Worker is about to be executed on the given Meadow...
220 221
@column submission_cmd_args ... these are the resource arguments (queue, memory,...) to give to the submission command
@column worker_cmd_args     ... and these are the arguments that are given to the worker command being submitted
222 223
*/

224
CREATE TABLE resource_description (
225 226
    resource_class_id       INTEGER      NOT NULL,
    meadow_type             VARCHAR(255) NOT NULL,
227 228
    submission_cmd_args     VARCHAR(255) NOT NULL DEFAULT '',
    worker_cmd_args         VARCHAR(255) NOT NULL DEFAULT '',
229 230 231 232 233

    PRIMARY KEY(resource_class_id, meadow_type)
) COLLATE=latin1_swedish_ci ENGINE=InnoDB;


234 235 236 237
/**
@header Job-related
@colour #1D73DA
*/
Jessica Severin's avatar
Jessica Severin committed
238

239 240
/**
@table  job
241

242
@colour #1D73DA
243

244 245 246 247
@desc The job is the heart of this system.  It is the kiosk or blackboard
    where workers find things to do and then post work for other works to do.
    These jobs are created prior to work being done, are claimed by workers,
    are updated as the work is done, with a final update on completion.
248

249
@column job_id                  autoincrement id
250
@column prev_job_id             previous job which created this one
251 252
@column analysis_id             the analysis_id needed to accomplish this job.
@column input_id                input data passed into Analysis:RunnableDB to control the work
253 254
@column param_id_stack          a CSV of job_ids whose input_ids contribute to the stack of local variables for the job
@column accu_id_stack           a CSV of job_ids whose accu's contribute to the stack of local variables for the job
255 256 257
@column worker_id               link to worker table to define which worker claimed this job
@column status                  state the job is in
@column retry_count             number times job had to be reset when worker failed to run it
Leo Gordon's avatar
Leo Gordon committed
258
@column completed               when the job was completed
259 260 261 262 263
@column runtime_msec            how long did it take to execute the job (or until the moment it failed)
@column query_count             how many SQL queries were run during this job
@column semaphore_count         if this count is >0, the job is conditionally blocked (until this count drops to 0 or below). Default=0 means "nothing is blocking me by default".
@column semaphored_job_id       the job_id of job S that is waiting for this job to decrease S's semaphore_count. Default=NULL means "I'm not blocking anything by default".
*/
264

265
CREATE TABLE job (
Leo Gordon's avatar
Leo Gordon committed
266 267 268 269
    job_id                  INTEGER     NOT NULL PRIMARY KEY AUTO_INCREMENT,
    prev_job_id             INTEGER              DEFAULT NULL,  -- the job that created this one using a dataflow rule
    analysis_id             INTEGER     NOT NULL,
    input_id                CHAR(255)   NOT NULL,
270 271
    param_id_stack          CHAR(64)    NOT NULL DEFAULT '',
    accu_id_stack           CHAR(64)    NOT NULL DEFAULT '',
Leo Gordon's avatar
Leo Gordon committed
272 273 274 275 276 277 278 279 280 281
    worker_id               INTEGER              DEFAULT NULL,
    status                  ENUM('SEMAPHORED','READY','CLAIMED','COMPILATION','PRE_CLEANUP','FETCH_INPUT','RUN','WRITE_OUTPUT','POST_CLEANUP','DONE','FAILED','PASSED_ON') DEFAULT 'READY' NOT NULL,
    retry_count             INTEGER     NOT NULL DEFAULT 0,
    completed               TIMESTAMP                    NULL,  -- mysql's special for "TIMESTAMP DEFAULT NULL"
    runtime_msec            INTEGER              DEFAULT NULL,
    query_count             INTEGER              DEFAULT NULL,

    semaphore_count         INTEGER     NOT NULL DEFAULT 0,
    semaphored_job_id       INTEGER              DEFAULT NULL,

282 283 284
    UNIQUE  KEY input_id_stacks_analysis    (input_id, param_id_stack, accu_id_stack, analysis_id), -- to avoid repeating tasks
            KEY analysis_status_retry       (analysis_id, status, retry_count),                     -- for claiming jobs
            KEY  worker_id                  (worker_id, status)                                     -- for fetching and releasing claimed jobs
285
) COLLATE=latin1_swedish_ci ENGINE=InnoDB;
286

287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305

/**
@table  job_file

@colour #1D73DA

@desc   For testing/debugging purposes both STDOUT and STDERR streams of each Job
        can be redirected into a separate log file.
        This table holds filesystem paths to one or both of those files.
        There is max one entry per job_id and retry.

@column job_id             foreign key
@column worker_id          link to worker table to define which worker claimed this job
@column retry              copy of retry_count of job as it was run
@column stdout_file        path to the job's STDOUT log
@column stderr_file        path to the job's STDERR log
*/

CREATE TABLE job_file (
Leo Gordon's avatar
Leo Gordon committed
306 307 308 309 310
    job_id                  INTEGER     NOT NULL,
    retry                   INTEGER     NOT NULL,
    worker_id               INTEGER     NOT NULL,
    stdout_file             VARCHAR(255),
    stderr_file             VARCHAR(255),
311 312 313

    PRIMARY KEY job_retry   (job_id, retry),
            KEY  worker_id  (worker_id)
Leo Gordon's avatar
Leo Gordon committed
314
) COLLATE=latin1_swedish_ci ENGINE=InnoDB;
Jessica Severin's avatar
Jessica Severin committed
315

Leo Gordon's avatar
Leo Gordon committed
316

317 318 319 320 321 322 323 324 325 326 327 328 329 330 331
/**
@table  accu

@colour #1D73DA

@desc   Accumulator for funneled dataflow.

@column sending_job_id     semaphoring job in the "box"
@column receiving_job_id   semaphored job outside the "box"
@column struct_name        name of the structured parameter
@column key_signature      locates the part of the structured parameter
@column value              value of the part
*/

CREATE TABLE accu (
Leo Gordon's avatar
Leo Gordon committed
332 333 334 335
    sending_job_id          INTEGER,
    receiving_job_id        INTEGER     NOT NULL,
    struct_name             VARCHAR(255) NOT NULL,
    key_signature           VARCHAR(255) NOT NULL,
336
    value                   TEXT,
Leo Gordon's avatar
Leo Gordon committed
337 338 339

    KEY accu_sending_idx (sending_job_id),
    KEY accu_receiving_idx (receiving_job_id)
340 341 342
) COLLATE=latin1_swedish_ci ENGINE=InnoDB;


343 344 345 346 347 348 349 350 351 352 353 354 355 356 357
/**
@table  analysis_data

@colour #1D73DA

@desc   A generic blob-storage hash.
        Currently the only legitimate use of this table is "overflow" of job.input_ids:
        when they grow longer than 254 characters the real data is stored in analysis_data instead,
        and the input_id contains the corresponding analysis_data_id.

@column analysis_data_id    primary id
@column data                text blob which holds the data
*/

CREATE TABLE analysis_data (
Leo Gordon's avatar
Leo Gordon committed
358 359
    analysis_data_id        INTEGER     NOT NULL PRIMARY KEY AUTO_INCREMENT,
    data                    TEXT,
360 361 362 363 364

            KEY (data(100))
) COLLATE=latin1_swedish_ci ENGINE=InnoDB;


365 366 367 368 369 370 371 372 373 374 375 376 377
/**
@table  hive_meta

@colour #000000

@desc This table keeps several important hive-specific pipeline-wide key-value pairs
        such as hive_sql_schema_version, hive_use_triggers and hive_pipeline_name.

@column meta_key        the KEY of KEY-VALUE pairs (primary key)
@column meta_value      the VALUE of KEY-VALUE pairs
*/

CREATE TABLE hive_meta (
378
    meta_key                VARCHAR(255) NOT NULL PRIMARY KEY,
379 380 381 382 383
    meta_value              TEXT

) COLLATE=latin1_swedish_ci ENGINE=MyISAM;


384 385 386 387 388 389
/**
@table  meta

@colour #000000

@desc This table comes from the Ensembl core schema.
Leo Gordon's avatar
Leo Gordon committed
390 391 392 393
        It is created here with the 'IF NOT EXISTS' option to avoid a potential clash
         if we are dealing with core-hive hybrid that is created in the wrong order.
        At the moment meta table is used
            (1) for compatibility with the Core API ('schema_version'),
394
            (2) to keep pipeline-wide parameters.
Leo Gordon's avatar
Leo Gordon committed
395 396 397 398 399

@column meta_id         auto-incrementing primary key, not really used per se
@column species_id      always 1, kept for compatibility with the Core API
@column meta_key        the KEY of KEY-VALUE pairs
@column meta_value      the VALUE of KEY-VALUE pairs
400 401 402
*/

CREATE TABLE IF NOT EXISTS meta (
403 404 405 406
    meta_id                 INTEGER      NOT NULL PRIMARY KEY AUTO_INCREMENT,
    species_id              INTEGER               DEFAULT 1,
    meta_key                VARCHAR(255) NOT NULL,
    meta_value              TEXT         NOT NULL,
407 408 409 410 411 412 413

    UNIQUE    KEY species_key_value_idx (species_id, meta_key, meta_value(255)),
              KEY species_value_idx (species_id, meta_value(255))

) COLLATE=latin1_swedish_ci ENGINE=MyISAM;


414 415 416 417 418 419 420 421 422 423 424
/**
@header worker table
@colour #24DA06
*/

/**
@table  worker

@colour #24DA06

@desc Entries of this table correspond to Worker objects of the API.
Leo Gordon's avatar
Leo Gordon committed
425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442
        Workers are created by inserting into this table
        so that there is only one instance of a Worker object in the database.
        As Workers live and do work, they update this table, and when they die they update again.

@column worker_id           unique ID of the Worker
@column meadow_type         type of the Meadow it is running on
@column meadow_name         name of the Meadow it is running on (for 'LOCAL' type is the same as host)
@column host                execution host name
@column process_id          identifies the Worker process on the Meadow (for 'LOCAL' is the OS PID)
@column resource_class_id   links to Worker's resource class
@column analysis_id         Analysis the Worker is specified into
@column work_done           how many jobs the Worker has completed successfully
@column status              current status of the Worker
@column born                when the Worker process was started
@column last_check_in       when the Worker last checked into the database
@column died                if defined, when the Worker died (or its premature death was first detected by GC)
@column cause_of_death      if defined, why did the Worker exit (or why it was killed)
@column log_dir             if defined, a filesystem directory where this Worker's output is logged
443 444 445
*/

CREATE TABLE worker (
Leo Gordon's avatar
Leo Gordon committed
446
    worker_id               INTEGER     NOT NULL PRIMARY KEY AUTO_INCREMENT,
447 448 449 450
    meadow_type             VARCHAR(255) NOT NULL,
    meadow_name             VARCHAR(255) NOT NULL,
    host                    VARCHAR(255) NOT NULL,
    process_id              VARCHAR(255) NOT NULL,
Leo Gordon's avatar
Leo Gordon committed
451 452 453 454 455 456 457 458 459 460
    resource_class_id       INTEGER              DEFAULT NULL,

    analysis_id             INTEGER              DEFAULT NULL,
    work_done               INTEGER     NOT NULL DEFAULT 0,
    status                  ENUM('SPECIALIZATION','COMPILATION','READY','PRE_CLEANUP','FETCH_INPUT','RUN','WRITE_OUTPUT','POST_CLEANUP','DEAD') DEFAULT 'READY' NOT NULL,
    born                    TIMESTAMP   NOT NULL DEFAULT CURRENT_TIMESTAMP,
    last_check_in           TIMESTAMP   NOT NULL,
    died                    TIMESTAMP                    NULL,  -- mysql's special for "TIMESTAMP DEFAULT NULL"
    cause_of_death          ENUM('NO_ROLE', 'NO_WORK', 'JOB_LIMIT', 'HIVE_OVERLOAD', 'LIFESPAN', 'CONTAMINATED', 'KILLED_BY_USER', 'MEMLIMIT', 'RUNLIMIT', 'SEE_MSG', 'UNKNOWN') DEFAULT NULL,
    log_dir                 VARCHAR(255)         DEFAULT NULL,
461

Leo Gordon's avatar
Leo Gordon committed
462
            KEY analysis_status (analysis_id, status)
463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491
) COLLATE=latin1_swedish_ci ENGINE=InnoDB;


/**
@header Logging and monitoring
@colour #F4D20C
*/

/**
@table  log_message

@colour #08DAD8

@desc   When a Job or a job-less Worker (job_id=NULL) throws a "die" message
        for any reason, the message is recorded in this table.
        It may or may not indicate that the job was unsuccessful via is_error flag.
        Also $self->warning("...") messages are recorded with is_error=0.

@column log_message_id  an autoincremented primary id of the message
@column         job_id  the id of the job that threw the message (or NULL if it was outside of a message)
@column      worker_id  the 'current' worker
@column           time  when the message was thrown
@column          retry  retry_count of the job when the message was thrown (or NULL if no job)
@column         status  of the job or worker when the message was thrown
@column            msg  string that contains the message
@column       is_error  binary flag
*/

CREATE TABLE log_message (
Leo Gordon's avatar
Leo Gordon committed
492 493
    log_message_id          INTEGER     NOT NULL PRIMARY KEY AUTO_INCREMENT,
    job_id                  INTEGER              DEFAULT NULL,
494
    worker_id               INTEGER              DEFAULT NULL,
Leo Gordon's avatar
Leo Gordon committed
495 496 497 498 499 500
    time                    TIMESTAMP   NOT NULL DEFAULT CURRENT_TIMESTAMP,
    retry                   INTEGER              DEFAULT NULL,
    status                  ENUM('UNKNOWN','SPECIALIZATION','COMPILATION','READY','PRE_CLEANUP','FETCH_INPUT','RUN','WRITE_OUTPUT','POST_CLEANUP','PASSED_ON') DEFAULT 'UNKNOWN',
    msg                     TEXT,
    is_error                SMALLINT,

501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541
            KEY worker_id     (worker_id),
            KEY job_id        (job_id)
) COLLATE=latin1_swedish_ci ENGINE=InnoDB;


/**
@table  analysis_stats_monitor

@colour #F4D20C

@desc   A regular timestamped snapshot of the analysis_stats table.

@column time                    when this snapshot was taken

@column analysis_id             foreign-keyed to the corresponding analysis_base entry
@column batch_size              how many jobs are claimed in one claiming operation before Worker starts executing them
@column hive_capacity           a reciprocal limiter on the number of Workers running at the same time (dependent on Workers of other Analyses)
@column status                  cached state of the Analysis

@column total_job_count         total number of Jobs of this Analysis
@column semaphored_job_count    number of Jobs of this Analysis that are in SEMAPHORED state
@column ready_job_count         number of Jobs of this Analysis that are in READY state
@column done_job_count          number of Jobs of this Analysis that are in DONE state
@column failed_job_count        number of Jobs of this Analysis that are in FAILED state

@column num_running_workers     number of running Workers of this Analysis
@column num_required_workers    extra number of Workers of this Analysis needed to execute all READY jobs

@column behaviour               whether hive_capacity is set or is dynamically calculated based on timers
@column input_capacity          used to compute hive_capacity in DYNAMIC mode
@column output_capacity         used to compute hive_capacity in DYNAMIC mode

@column avg_msec_per_job        weighted average used to compute DYNAMIC hive_capacity
@column avg_input_msec_per_job  weighted average used to compute DYNAMIC hive_capacity
@column avg_run_msec_per_job    weighted average used to compute DYNAMIC hive_capacity
@column avg_output_msec_per_job weighted average used to compute DYNAMIC hive_capacity

@column last_update             when this entry was last updated
@column sync_lock               a binary lock flag to prevent simultaneous updates
*/

542
CREATE TABLE analysis_stats_monitor (
Leo Gordon's avatar
Leo Gordon committed
543
    time                    TIMESTAMP   NOT NULL DEFAULT CURRENT_TIMESTAMP,
544

Leo Gordon's avatar
Leo Gordon committed
545 546 547 548
    analysis_id             INTEGER     NOT NULL,
    batch_size              INTEGER     NOT NULL DEFAULT 1,
    hive_capacity           INTEGER              DEFAULT NULL,
    status                  ENUM('BLOCKED', 'LOADING', 'SYNCHING', 'EMPTY', 'READY', 'WORKING', 'ALL_CLAIMED', 'DONE', 'FAILED') DEFAULT 'EMPTY' NOT NULL,
549

Leo Gordon's avatar
Leo Gordon committed
550 551 552 553 554 555 556
    total_job_count         INTEGER     NOT NULL DEFAULT 0,
    semaphored_job_count    INTEGER     NOT NULL DEFAULT 0,
    ready_job_count         INTEGER     NOT NULL DEFAULT 0,
    done_job_count          INTEGER     NOT NULL DEFAULT 0,
    failed_job_count        INTEGER     NOT NULL DEFAULT 0,
    num_running_workers     INTEGER     NOT NULL DEFAULT 0,
    num_required_workers    INTEGER     NOT NULL DEFAULT 0,
557

Leo Gordon's avatar
Leo Gordon committed
558 559 560
    behaviour               ENUM('STATIC', 'DYNAMIC') NOT NULL DEFAULT 'STATIC',
    input_capacity          INTEGER     NOT NULL DEFAULT 4,
    output_capacity         INTEGER     NOT NULL DEFAULT 4,
561

Leo Gordon's avatar
Leo Gordon committed
562 563 564 565
    avg_msec_per_job        INTEGER              DEFAULT NULL,
    avg_input_msec_per_job  INTEGER              DEFAULT NULL,
    avg_run_msec_per_job    INTEGER              DEFAULT NULL,
    avg_output_msec_per_job INTEGER              DEFAULT NULL,
566

Leo Gordon's avatar
Leo Gordon committed
567 568
    last_update             TIMESTAMP                    NULL,  -- mysql's special for "TIMESTAMP DEFAULT NULL"
    sync_lock               SMALLINT    NOT NULL DEFAULT 0
Leo Gordon's avatar
Leo Gordon committed
569

Leo Gordon's avatar
Leo Gordon committed
570
) COLLATE=latin1_swedish_ci ENGINE=InnoDB;
571

572

573 574 575 576 577 578 579 580 581 582 583 584 585 586
/**
@table  monitor

@colour #F4D20C

@desc   A regular collated snapshot of the Worker table.

@column time            when this snapshot was taken
@column workers         number of running workers
@column throughput      average numb of completed Jobs per sec. of the hive (this number is calculated using running workers only)
@column per_worker      average numb of completed Jobs per sec. per Worker (this number is calculated using running workers only)
@column analysis        a comma-separated list of analyses running at the time of snapshot

*/
587 588

CREATE TABLE monitor (
Leo Gordon's avatar
Leo Gordon committed
589 590 591 592
    time                    TIMESTAMP   NOT NULL DEFAULT CURRENT_TIMESTAMP,
    workers                 INTEGER     NOT NULL DEFAULT 0,
    throughput              FLOAT                DEFAULT NULL,
    per_worker              FLOAT                DEFAULT NULL,
593
    analysis                TEXT                 DEFAULT NULL   -- not just one, but a list of logic_names
594

Leo Gordon's avatar
Leo Gordon committed
595
) COLLATE=latin1_swedish_ci ENGINE=InnoDB;
596 597