1 2 3 4 5 6 7 8 9 10 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 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219
|
/*-------------------------------------------------------------------------
*
* connection.c
* Connection management functions for postgres_fdw
*
* Portions Copyright (c) 2012-2018, PostgreSQL Global Development Group
*
* IDENTIFICATION
* contrib/postgres_fdw/connection.c
*
*-------------------------------------------------------------------------
*/
#include "postgres.h"
#include "postgres_fdw.h"
#include "access/htup_details.h"
#include "catalog/pg_user_mapping.h"
#include "access/xact.h"
#include "mb/pg_wchar.h"
#include "miscadmin.h"
#include "pgstat.h"
#include "storage/latch.h"
#include "utils/hsearch.h"
#include "utils/inval.h"
#include "utils/memutils.h"
#include "utils/syscache.h"
/*
* Connection cache hash table entry
*
* The lookup key in this hash table is the user mapping OID. We use just one
* connection per user mapping ID, which ensures that all the scans use the
* same snapshot during a query. Using the user mapping OID rather than
* the foreign server OID + user OID avoids creating multiple connections when
* the public user mapping applies to all user OIDs.
*
* The "conn" pointer can be NULL if we don't currently have a live connection.
* When we do have a connection, xact_depth tracks the current depth of
* transactions and subtransactions open on the remote side. We need to issue
* commands at the same nesting depth on the remote as we're executing at
* ourselves, so that rolling back a subtransaction will kill the right
* queries and not the wrong ones.
*/
typedef Oid ConnCacheKey;
typedef struct ConnCacheEntry
{
ConnCacheKey key; /* hash key (must be first) */
PGconn *conn; /* connection to foreign server, or NULL */
/* Remaining fields are invalid when conn is NULL: */
int xact_depth; /* 0 = no xact open, 1 = main xact open, 2 =
* one level of subxact open, etc */
bool have_prep_stmt; /* have we prepared any stmts in this xact? */
bool have_error; /* have any subxacts aborted in this xact? */
bool changing_xact_state; /* xact state change in process */
bool invalidated; /* true if reconnect is pending */
uint32 server_hashvalue; /* hash value of foreign server OID */
uint32 mapping_hashvalue; /* hash value of user mapping OID */
} ConnCacheEntry;
/*
* Connection cache (initialized on first use)
*/
static HTAB *ConnectionHash = NULL;
/* for assigning cursor numbers and prepared statement numbers */
static unsigned int cursor_number = 0;
static unsigned int prep_stmt_number = 0;
/* tracks whether any work is needed in callback functions */
static bool xact_got_connection = false;
/* prototypes of private functions */
static PGconn *connect_pg_server(ForeignServer *server, UserMapping *user);
static void disconnect_pg_server(ConnCacheEntry *entry);
static void check_conn_params(const char **keywords, const char **values, UserMapping *user);
static void configure_remote_session(PGconn *conn);
static void do_sql_command(PGconn *conn, const char *sql);
static void begin_remote_xact(ConnCacheEntry *entry);
static void pgfdw_xact_callback(XactEvent event, void *arg);
static void pgfdw_subxact_callback(SubXactEvent event,
SubTransactionId mySubid,
SubTransactionId parentSubid,
void *arg);
static void pgfdw_inval_callback(Datum arg, int cacheid, uint32 hashvalue);
static void pgfdw_reject_incomplete_xact_state_change(ConnCacheEntry *entry);
static bool pgfdw_cancel_query(PGconn *conn);
static bool pgfdw_exec_cleanup_query(PGconn *conn, const char *query,
bool ignore_errors);
static bool pgfdw_get_cleanup_result(PGconn *conn, TimestampTz endtime,
PGresult **result);
/*
* Get a PGconn which can be used to execute queries on the remote PostgreSQL
* server with the user's authorization. A new connection is established
* if we don't already have a suitable one, and a transaction is opened at
* the right subtransaction nesting depth if we didn't do that already.
*
* will_prep_stmt must be true if caller intends to create any prepared
* statements. Since those don't go away automatically at transaction end
* (not even on error), we need this flag to cue manual cleanup.
*/
PGconn *
GetConnection(UserMapping *user, bool will_prep_stmt)
{
bool found;
ConnCacheEntry *entry;
ConnCacheKey key;
/* First time through, initialize connection cache hashtable */
if (ConnectionHash == NULL)
{
HASHCTL ctl;
MemSet(&ctl, 0, sizeof(ctl));
ctl.keysize = sizeof(ConnCacheKey);
ctl.entrysize = sizeof(ConnCacheEntry);
/* allocate ConnectionHash in the cache context */
ctl.hcxt = CacheMemoryContext;
ConnectionHash = hash_create("postgres_fdw connections", 8,
&ctl,
HASH_ELEM | HASH_BLOBS | HASH_CONTEXT);
/*
* Register some callback functions that manage connection cleanup.
* This should be done just once in each backend.
*/
RegisterXactCallback(pgfdw_xact_callback, NULL);
RegisterSubXactCallback(pgfdw_subxact_callback, NULL);
CacheRegisterSyscacheCallback(FOREIGNSERVEROID,
pgfdw_inval_callback, (Datum) 0);
CacheRegisterSyscacheCallback(USERMAPPINGOID,
pgfdw_inval_callback, (Datum) 0);
}
/* Set flag that we did GetConnection during the current transaction */
xact_got_connection = true;
/* Create hash key for the entry. Assume no pad bytes in key struct */
key = user->umid;
/*
* Find or create cached entry for requested connection.
*/
entry = hash_search(ConnectionHash, &key, HASH_ENTER, &found);
if (!found)
{
/*
* We need only clear "conn" here; remaining fields will be filled
* later when "conn" is set.
*/
entry->conn = NULL;
}
/* Reject further use of connections which failed abort cleanup. */
pgfdw_reject_incomplete_xact_state_change(entry);
/*
* If the connection needs to be remade due to invalidation, disconnect as
* soon as we're out of all transactions.
*/
if (entry->conn != NULL && entry->invalidated && entry->xact_depth == 0)
{
elog(DEBUG3, "closing connection %p for option changes to take effect",
entry->conn);
disconnect_pg_server(entry);
}
/*
* We don't check the health of cached connection here, because it would
* require some overhead. Broken connection will be detected when the
* connection is actually used.
*/
/*
* If cache entry doesn't have a connection, we have to establish a new
* connection. (If connect_pg_server throws an error, the cache entry
* will remain in a valid empty state, ie conn == NULL.)
*/
if (entry->conn == NULL)
{
ForeignServer *server = GetForeignServer(user->serverid);
/* Reset all transient state fields, to be sure all are clean */
entry->xact_depth = 0;
entry->have_prep_stmt = false;
entry->have_error = false;
entry->changing_xact_state = false;
entry->invalidated = false;
entry->server_hashvalue =
GetSysCacheHashValue1(FOREIGNSERVEROID,
ObjectIdGetDatum(server->serverid));
entry->mapping_hashvalue =
GetSysCacheHashValue1(USERMAPPINGOID,
ObjectIdGetDatum(user->umid));
/* Now try to make the connection */
entry->conn = connect_pg_server(server, user);
elog(DEBUG3, "new postgres_fdw connection %p for server \"%s\" (user mapping oid %u, userid %u)",
entry->conn, server->servername, user->umid, user->userid);
}
/*
* Start a new transaction or subtransaction if needed.
*/
begin_remote_xact(entry);
/* Remember if caller will prepare statements */
entry->have_prep_stmt |= will_prep_stmt;
return entry->conn;
}
/*
* Connect to remote server using specified server and user mapping properties.
*/
static PGconn *
connect_pg_server(ForeignServer *server, UserMapping *user)
{
PGconn *volatile conn = NULL;
/*
* Use PG_TRY block to ensure closing connection on error.
*/
PG_TRY();
{
const char **keywords;
const char **values;
int n;
/*
* Construct connection params from generic options of ForeignServer
* and UserMapping. (Some of them might not be libpq options, in
* which case we'll just waste a few array slots.) Add 3 extra slots
* for fallback_application_name, client_encoding, end marker.
*/
n = list_length(server->options) + list_length(user->options) + 3;
keywords = (const char **) palloc(n * sizeof(char *));
values = (const char **) palloc(n * sizeof(char *));
n = 0;
n += ExtractConnectionOptions(server->options,
keywords + n, values + n);
n += ExtractConnectionOptions(user->options,
keywords + n, values + n);
/* Use "postgres_fdw" as fallback_application_name. */
keywords[n] = "fallback_application_name";
values[n] = "postgres_fdw";
n++;
/* Set client_encoding so that libpq can convert encoding properly. */
keywords[n] = "client_encoding";
values[n] = GetDatabaseEncodingName();
n++;
keywords[n] = values[n] = NULL;
/* verify connection parameters and make connection */
check_conn_params(keywords, values, user);
conn = PQconnectdbParams(keywords, values, false);
if (!conn || PQstatus(conn) != CONNECTION_OK)
ereport(ERROR,
(errcode(ERRCODE_SQLCLIENT_UNABLE_TO_ESTABLISH_SQLCONNECTION),
errmsg("could not connect to server \"%s\"",
server->servername),
errdetail_internal("%s", pchomp(PQerrorMessage(conn)))));
/*
* Check that non-superuser has used password to establish connection;
* otherwise, he's piggybacking on the postgres server's user
* identity. See also dblink_security_check() in contrib/dblink.
*/
if (!superuser_arg(user->userid) && !PQconnectionUsedPassword(conn))
ereport(ERROR,
(errcode(ERRCODE_S_R_E_PROHIBITED_SQL_STATEMENT_ATTEMPTED),
errmsg("password is required"),
errdetail("Non-superuser cannot connect if the server does not request a password."),
errhint("Target server's authentication method must be changed.")));
/* Prepare new session for use */
configure_remote_session(conn);
pfree(keywords);
pfree(values);
}
PG_CATCH();
{
/* Release PGconn data structure if we managed to create one */
if (conn)
PQfinish(conn);
PG_RE_THROW();
}
PG_END_TRY();
return conn;
}
/*
* Disconnect any open connection for a connection cache entry.
*/
static void
disconnect_pg_server(ConnCacheEntry *entry)
{
if (entry->conn != NULL)
{
PQfinish(entry->conn);
entry->conn = NULL;
}
}
/*
* For non-superusers, insist that the connstr specify a password. This
* prevents a password from being picked up from .pgpass, a service file,
* the environment, etc. We don't want the postgres user's passwords
* to be accessible to non-superusers. (See also dblink_connstr_check in
* contrib/dblink.)
*/
static void
check_conn_params(const char **keywords, const char **values, UserMapping *user)
{
int i;
/* no check required if superuser */
if (superuser_arg(user->userid))
return;
/* ok if params contain a non-empty password */
for (i = 0; keywords[i] != NULL; i++)
{
if (strcmp(keywords[i], "password") == 0 && values[i][0] != '\0')
return;
}
ereport(ERROR,
(errcode(ERRCODE_S_R_E_PROHIBITED_SQL_STATEMENT_ATTEMPTED),
errmsg("password is required"),
errdetail("Non-superusers must provide a password in the user mapping.")));
}
/*
* Issue SET commands to make sure remote session is configured properly.
*
* We do this just once at connection, assuming nothing will change the
* values later. Since we'll never send volatile function calls to the
* remote, there shouldn't be any way to break this assumption from our end.
* It's possible to think of ways to break it at the remote end, eg making
* a foreign table point to a view that includes a set_config call ---
* but once you admit the possibility of a malicious view definition,
* there are any number of ways to break things.
*/
static void
configure_remote_session(PGconn *conn)
{
int remoteversion = PQserverVersion(conn);
/* Force the search path to contain only pg_catalog (see deparse.c) */
do_sql_command(conn, "SET search_path = pg_catalog");
/*
* Set remote timezone; this is basically just cosmetic, since all
* transmitted and returned timestamptzs should specify a zone explicitly
* anyway. However it makes the regression test outputs more predictable.
*
* We don't risk setting remote zone equal to ours, since the remote
* server might use a different timezone database. Instead, use UTC
* (quoted, because very old servers are picky about case).
*/
do_sql_command(conn, "SET timezone = 'UTC'");
/*
* Set values needed to ensure unambiguous data output from remote. (This
* logic should match what pg_dump does. See also set_transmission_modes
* in postgres_fdw.c.)
*/
do_sql_command(conn, "SET datestyle = ISO");
if (remoteversion >= 80400)
do_sql_command(conn, "SET intervalstyle = postgres");
if (remoteversion >= 90000)
do_sql_command(conn, "SET extra_float_digits = 3");
else
do_sql_command(conn, "SET extra_float_digits = 2");
}
/*
* Convenience subroutine to issue a non-data-returning SQL command to remote
*/
static void
do_sql_command(PGconn *conn, const char *sql)
{
PGresult *res;
if (!PQsendQuery(conn, sql))
pgfdw_report_error(ERROR, NULL, conn, false, sql);
res = pgfdw_get_result(conn, sql);
if (PQresultStatus(res) != PGRES_COMMAND_OK)
pgfdw_report_error(ERROR, res, conn, true, sql);
PQclear(res);
}
/*
* Start remote transaction or subtransaction, if needed.
*
* Note that we always use at least REPEATABLE READ in the remote session.
* This is so that, if a query initiates multiple scans of the same or
* different foreign tables, we will get snapshot-consistent results from
* those scans. A disadvantage is that we can't provide sane emulation of
* READ COMMITTED behavior --- it would be nice if we had some other way to
* control which remote queries share a snapshot.
*/
static void
begin_remote_xact(ConnCacheEntry *entry)
{
int curlevel = GetCurrentTransactionNestLevel();
/* Start main transaction if we haven't yet */
if (entry->xact_depth <= 0)
{
const char *sql;
elog(DEBUG3, "starting remote transaction on connection %p",
entry->conn);
if (IsolationIsSerializable())
sql = "START TRANSACTION ISOLATION LEVEL SERIALIZABLE";
else
sql = "START TRANSACTION ISOLATION LEVEL REPEATABLE READ";
entry->changing_xact_state = true;
do_sql_command(entry->conn, sql);
entry->xact_depth = 1;
entry->changing_xact_state = false;
}
/*
* If we're in a subtransaction, stack up savepoints to match our level.
* This ensures we can rollback just the desired effects when a
* subtransaction aborts.
*/
while (entry->xact_depth < curlevel)
{
char sql[64];
snprintf(sql, sizeof(sql), "SAVEPOINT s%d", entry->xact_depth + 1);
entry->changing_xact_state = true;
do_sql_command(entry->conn, sql);
entry->xact_depth++;
entry->changing_xact_state = false;
}
}
/*
* Release connection reference count created by calling GetConnection.
*/
void
ReleaseConnection(PGconn *conn)
{
/*
* Currently, we don't actually track connection references because all
* cleanup is managed on a transaction or subtransaction basis instead. So
* there's nothing to do here.
*/
}
/*
* Assign a "unique" number for a cursor.
*
* These really only need to be unique per connection within a transaction.
* For the moment we ignore the per-connection point and assign them across
* all connections in the transaction, but we ask for the connection to be
* supplied in case we want to refine that.
*
* Note that even if wraparound happens in a very long transaction, actual
* collisions are highly improbable; just be sure to use %u not %d to print.
*/
unsigned int
GetCursorNumber(PGconn *conn)
{
return ++cursor_number;
}
/*
* Assign a "unique" number for a prepared statement.
*
* This works much like GetCursorNumber, except that we never reset the counter
* within a session. That's because we can't be 100% sure we've gotten rid
* of all prepared statements on all connections, and it's not really worth
* increasing the risk of prepared-statement name collisions by resetting.
*/
unsigned int
GetPrepStmtNumber(PGconn *conn)
{
return ++prep_stmt_number;
}
/*
* Submit a query and wait for the result.
*
* This function is interruptible by signals.
*
* Caller is responsible for the error handling on the result.
*/
PGresult *
pgfdw_exec_query(PGconn *conn, const char *query)
{
/*
* Submit a query. Since we don't use non-blocking mode, this also can
* block. But its risk is relatively small, so we ignore that for now.
*/
if (!PQsendQuery(conn, query))
pgfdw_report_error(ERROR, NULL, conn, false, query);
/* Wait for the result. */
return pgfdw_get_result(conn, query);
}
/*
* Wait for the result from a prior asynchronous execution function call.
*
* This function offers quick responsiveness by checking for any interruptions.
*
* This function emulates PQexec()'s behavior of returning the last result
* when there are many.
*
* Caller is responsible for the error handling on the result.
*/
PGresult *
pgfdw_get_result(PGconn *conn, const char *query)
{
PGresult *volatile last_res = NULL;
/* In what follows, do not leak any PGresults on an error. */
PG_TRY();
{
for (;;)
{
PGresult *res;
while (PQisBusy(conn))
{
int wc;
/* Sleep until there's something to do */
wc = WaitLatchOrSocket(MyLatch,
WL_LATCH_SET | WL_SOCKET_READABLE,
PQsocket(conn),
-1L, PG_WAIT_EXTENSION);
ResetLatch(MyLatch);
CHECK_FOR_INTERRUPTS();
/* Data available in socket? */
if (wc & WL_SOCKET_READABLE)
{
if (!PQconsumeInput(conn))
pgfdw_report_error(ERROR, NULL, conn, false, query);
}
}
res = PQgetResult(conn);
if (res == NULL)
break; /* query is complete */
PQclear(last_res);
last_res = res;
}
}
PG_CATCH();
{
PQclear(last_res);
PG_RE_THROW();
}
PG_END_TRY();
return last_res;
}
/*
* Report an error we got from the remote server.
*
* elevel: error level to use (typically ERROR, but might be less)
* res: PGresult containing the error
* conn: connection we did the query on
* clear: if true, PQclear the result (otherwise caller will handle it)
* sql: NULL, or text of remote command we tried to execute
*
* Note: callers that choose not to throw ERROR for a remote error are
* responsible for making sure that the associated ConnCacheEntry gets
* marked with have_error = true.
*/
void
pgfdw_report_error(int elevel, PGresult *res, PGconn *conn,
bool clear, const char *sql)
{
/* If requested, PGresult must be released before leaving this function. */
PG_TRY();
{
char *diag_sqlstate = PQresultErrorField(res, PG_DIAG_SQLSTATE);
char *message_primary = PQresultErrorField(res, PG_DIAG_MESSAGE_PRIMARY);
char *message_detail = PQresultErrorField(res, PG_DIAG_MESSAGE_DETAIL);
char *message_hint = PQresultErrorField(res, PG_DIAG_MESSAGE_HINT);
char *message_context = PQresultErrorField(res, PG_DIAG_CONTEXT);
int sqlstate;
if (diag_sqlstate)
sqlstate = MAKE_SQLSTATE(diag_sqlstate[0],
diag_sqlstate[1],
diag_sqlstate[2],
diag_sqlstate[3],
diag_sqlstate[4]);
else
sqlstate = ERRCODE_CONNECTION_FAILURE;
/*
* If we don't get a message from the PGresult, try the PGconn. This
* is needed because for connection-level failures, PQexec may just
* return NULL, not a PGresult at all.
*/
if (message_primary == NULL)
message_primary = pchomp(PQerrorMessage(conn));
ereport(elevel,
(errcode(sqlstate),
(message_primary != NULL && message_primary[0] != '\0') ?
errmsg_internal("%s", message_primary) :
errmsg("could not obtain message string for remote error"),
message_detail ? errdetail_internal("%s", message_detail) : 0,
message_hint ? errhint("%s", message_hint) : 0,
message_context ? errcontext("%s", message_context) : 0,
sql ? errcontext("remote SQL command: %s", sql) : 0));
}
PG_CATCH();
{
if (clear)
PQclear(res);
PG_RE_THROW();
}
PG_END_TRY();
if (clear)
PQclear(res);
}
/*
* pgfdw_xact_callback --- cleanup at main-transaction end.
*
* This runs just late enough that it must not enter user-defined code
* locally. (Entering such code on the remote side is fine. Its remote
* COMMIT TRANSACTION may run deferred triggers.)
*/
static void
pgfdw_xact_callback(XactEvent event, void *arg)
{
HASH_SEQ_STATUS scan;
ConnCacheEntry *entry;
/* Quick exit if no connections were touched in this transaction. */
if (!xact_got_connection)
return;
/*
* Scan all connection cache entries to find open remote transactions, and
* close them.
*/
hash_seq_init(&scan, ConnectionHash);
while ((entry = (ConnCacheEntry *) hash_seq_search(&scan)))
{
PGresult *res;
/* Ignore cache entry if no open connection right now */
if (entry->conn == NULL)
continue;
/* If it has an open remote transaction, try to close it */
if (entry->xact_depth > 0)
{
bool abort_cleanup_failure = false;
elog(DEBUG3, "closing remote transaction on connection %p",
entry->conn);
switch (event)
{
case XACT_EVENT_PARALLEL_PRE_COMMIT:
case XACT_EVENT_PRE_COMMIT:
/*
* If abort cleanup previously failed for this connection,
* we can't issue any more commands against it.
*/
pgfdw_reject_incomplete_xact_state_change(entry);
/* Commit all remote transactions during pre-commit */
entry->changing_xact_state = true;
do_sql_command(entry->conn, "COMMIT TRANSACTION");
entry->changing_xact_state = false;
/*
* If there were any errors in subtransactions, and we
* made prepared statements, do a DEALLOCATE ALL to make
* sure we get rid of all prepared statements. This is
* annoying and not terribly bulletproof, but it's
* probably not worth trying harder.
*
* DEALLOCATE ALL only exists in 8.3 and later, so this
* constrains how old a server postgres_fdw can
* communicate with. We intentionally ignore errors in
* the DEALLOCATE, so that we can hobble along to some
* extent with older servers (leaking prepared statements
* as we go; but we don't really support update operations
* pre-8.3 anyway).
*/
if (entry->have_prep_stmt && entry->have_error)
{
res = PQexec(entry->conn, "DEALLOCATE ALL");
PQclear(res);
}
entry->have_prep_stmt = false;
entry->have_error = false;
break;
case XACT_EVENT_PRE_PREPARE:
/*
* We disallow any remote transactions, since it's not
* very reasonable to hold them open until the prepared
* transaction is committed. For the moment, throw error
* unconditionally; later we might allow read-only cases.
* Note that the error will cause us to come right back
* here with event == XACT_EVENT_ABORT, so we'll clean up
* the connection state at that point.
*/
ereport(ERROR,
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
errmsg("cannot PREPARE a transaction that has operated on postgres_fdw foreign tables")));
break;
case XACT_EVENT_PARALLEL_COMMIT:
case XACT_EVENT_COMMIT:
case XACT_EVENT_PREPARE:
/* Pre-commit should have closed the open transaction */
elog(ERROR, "missed cleaning up connection during pre-commit");
break;
case XACT_EVENT_PARALLEL_ABORT:
case XACT_EVENT_ABORT:
/*
* Don't try to clean up the connection if we're already
* in error recursion trouble.
*/
if (in_error_recursion_trouble())
entry->changing_xact_state = true;
/*
* If connection is already unsalvageable, don't touch it
* further.
*/
if (entry->changing_xact_state)
break;
/*
* Mark this connection as in the process of changing
* transaction state.
*/
entry->changing_xact_state = true;
/* Assume we might have lost track of prepared statements */
entry->have_error = true;
/*
* If a command has been submitted to the remote server by
* using an asynchronous execution function, the command
* might not have yet completed. Check to see if a
* command is still being processed by the remote server,
* and if so, request cancellation of the command.
*/
if (PQtransactionStatus(entry->conn) == PQTRANS_ACTIVE &&
!pgfdw_cancel_query(entry->conn))
{
/* Unable to cancel running query. */
abort_cleanup_failure = true;
}
else if (!pgfdw_exec_cleanup_query(entry->conn,
"ABORT TRANSACTION",
false))
{
/* Unable to abort remote transaction. */
abort_cleanup_failure = true;
}
else if (entry->have_prep_stmt && entry->have_error &&
!pgfdw_exec_cleanup_query(entry->conn,
"DEALLOCATE ALL",
true))
{
/* Trouble clearing prepared statements. */
abort_cleanup_failure = true;
}
else
{
entry->have_prep_stmt = false;
entry->have_error = false;
}
/* Disarm changing_xact_state if it all worked. */
entry->changing_xact_state = abort_cleanup_failure;
break;
}
}
/* Reset state to show we're out of a transaction */
entry->xact_depth = 0;
/*
* If the connection isn't in a good idle state or it is marked as
* invalid, then discard it to recover. Next GetConnection will open a
* new connection.
*/
if (PQstatus(entry->conn) != CONNECTION_OK ||
PQtransactionStatus(entry->conn) != PQTRANS_IDLE ||
entry->changing_xact_state ||
entry->invalidated)
{
elog(DEBUG3, "discarding connection %p", entry->conn);
disconnect_pg_server(entry);
}
}
/*
* Regardless of the event type, we can now mark ourselves as out of the
* transaction. (Note: if we are here during PRE_COMMIT or PRE_PREPARE,
* this saves a useless scan of the hashtable during COMMIT or PREPARE.)
*/
xact_got_connection = false;
/* Also reset cursor numbering for next transaction */
cursor_number = 0;
}
/*
* pgfdw_subxact_callback --- cleanup at subtransaction end.
*/
static void
pgfdw_subxact_callback(SubXactEvent event, SubTransactionId mySubid,
SubTransactionId parentSubid, void *arg)
{
HASH_SEQ_STATUS scan;
ConnCacheEntry *entry;
int curlevel;
/* Nothing to do at subxact start, nor after commit. */
if (!(event == SUBXACT_EVENT_PRE_COMMIT_SUB ||
event == SUBXACT_EVENT_ABORT_SUB))
return;
/* Quick exit if no connections were touched in this transaction. */
if (!xact_got_connection)
return;
/*
* Scan all connection cache entries to find open remote subtransactions
* of the current level, and close them.
*/
curlevel = GetCurrentTransactionNestLevel();
hash_seq_init(&scan, ConnectionHash);
while ((entry = (ConnCacheEntry *) hash_seq_search(&scan)))
{
char sql[100];
/*
* We only care about connections with open remote subtransactions of
* the current level.
*/
if (entry->conn == NULL || entry->xact_depth < curlevel)
continue;
if (entry->xact_depth > curlevel)
elog(ERROR, "missed cleaning up remote subtransaction at level %d",
entry->xact_depth);
if (event == SUBXACT_EVENT_PRE_COMMIT_SUB)
{
/*
* If abort cleanup previously failed for this connection, we
* can't issue any more commands against it.
*/
pgfdw_reject_incomplete_xact_state_change(entry);
/* Commit all remote subtransactions during pre-commit */
snprintf(sql, sizeof(sql), "RELEASE SAVEPOINT s%d", curlevel);
entry->changing_xact_state = true;
do_sql_command(entry->conn, sql);
entry->changing_xact_state = false;
}
else if (in_error_recursion_trouble())
{
/*
* Don't try to clean up the connection if we're already in error
* recursion trouble.
*/
entry->changing_xact_state = true;
}
else if (!entry->changing_xact_state)
{
bool abort_cleanup_failure = false;
/* Remember that abort cleanup is in progress. */
entry->changing_xact_state = true;
/* Assume we might have lost track of prepared statements */
entry->have_error = true;
/*
* If a command has been submitted to the remote server by using
* an asynchronous execution function, the command might not have
* yet completed. Check to see if a command is still being
* processed by the remote server, and if so, request cancellation
* of the command.
*/
if (PQtransactionStatus(entry->conn) == PQTRANS_ACTIVE &&
!pgfdw_cancel_query(entry->conn))
abort_cleanup_failure = true;
else
{
/* Rollback all remote subtransactions during abort */
snprintf(sql, sizeof(sql),
"ROLLBACK TO SAVEPOINT s%d; RELEASE SAVEPOINT s%d",
curlevel, curlevel);
if (!pgfdw_exec_cleanup_query(entry->conn, sql, false))
abort_cleanup_failure = true;
}
/* Disarm changing_xact_state if it all worked. */
entry->changing_xact_state = abort_cleanup_failure;
}
/* OK, we're outta that level of subtransaction */
entry->xact_depth--;
}
}
/*
* Connection invalidation callback function
*
* After a change to a pg_foreign_server or pg_user_mapping catalog entry,
* close connections depending on that entry immediately if current transaction
* has not used those connections yet. Otherwise, mark those connections as
* invalid and then make pgfdw_xact_callback() close them at the end of current
* transaction, since they cannot be closed in the midst of the transaction
* using them. Closed connections will be remade at the next opportunity if
* necessary.
*
* Although most cache invalidation callbacks blow away all the related stuff
* regardless of the given hashvalue, connections are expensive enough that
* it's worth trying to avoid that.
*
* NB: We could avoid unnecessary disconnection more strictly by examining
* individual option values, but it seems too much effort for the gain.
*/
static void
pgfdw_inval_callback(Datum arg, int cacheid, uint32 hashvalue)
{
HASH_SEQ_STATUS scan;
ConnCacheEntry *entry;
Assert(cacheid == FOREIGNSERVEROID || cacheid == USERMAPPINGOID);
/* ConnectionHash must exist already, if we're registered */
hash_seq_init(&scan, ConnectionHash);
while ((entry = (ConnCacheEntry *) hash_seq_search(&scan)))
{
/* Ignore invalid entries */
if (entry->conn == NULL)
continue;
/* hashvalue == 0 means a cache reset, must clear all state */
if (hashvalue == 0 ||
(cacheid == FOREIGNSERVEROID &&
entry->server_hashvalue == hashvalue) ||
(cacheid == USERMAPPINGOID &&
entry->mapping_hashvalue == hashvalue))
{
/*
* Close the connection immediately if it's not used yet in this
* transaction. Otherwise mark it as invalid so that
* pgfdw_xact_callback() can close it at the end of this
* transaction.
*/
if (entry->xact_depth == 0)
{
elog(DEBUG3, "discarding connection %p", entry->conn);
disconnect_pg_server(entry);
}
else
entry->invalidated = true;
}
}
}
/*
* Raise an error if the given connection cache entry is marked as being
* in the middle of an xact state change. This should be called at which no
* such change is expected to be in progress; if one is found to be in
* progress, it means that we aborted in the middle of a previous state change
* and now don't know what the remote transaction state actually is.
* Such connections can't safely be further used. Re-establishing the
* connection would change the snapshot and roll back any writes already
* performed, so that's not an option, either. Thus, we must abort.
*/
static void
pgfdw_reject_incomplete_xact_state_change(ConnCacheEntry *entry)
{
HeapTuple tup;
Form_pg_user_mapping umform;
ForeignServer *server;
/* nothing to do for inactive entries and entries of sane state */
if (entry->conn == NULL || !entry->changing_xact_state)
return;
/* make sure this entry is inactive */
disconnect_pg_server(entry);
/* find server name to be shown in the message below */
tup = SearchSysCache1(USERMAPPINGOID,
ObjectIdGetDatum(entry->key));
if (!HeapTupleIsValid(tup))
elog(ERROR, "cache lookup failed for user mapping %u", entry->key);
umform = (Form_pg_user_mapping) GETSTRUCT(tup);
server = GetForeignServer(umform->umserver);
ReleaseSysCache(tup);
ereport(ERROR,
(errcode(ERRCODE_CONNECTION_EXCEPTION),
errmsg("connection to server \"%s\" was lost",
server->servername)));
}
/*
* Cancel the currently-in-progress query (whose query text we do not have)
* and ignore the result. Returns true if we successfully cancel the query
* and discard any pending result, and false if not.
*
* It's not a huge problem if we throw an ERROR here, but if we get into error
* recursion trouble, we'll end up slamming the connection shut, which will
* necessitate failing the entire toplevel transaction even if subtransactions
* were used. Try to use WARNING where we can.
*/
static bool
pgfdw_cancel_query(PGconn *conn)
{
PGcancel *cancel;
char errbuf[256];
PGresult *result = NULL;
TimestampTz endtime;
/*
* If it takes too long to cancel the query and discard the result, assume
* the connection is dead.
*/
endtime = TimestampTzPlusMilliseconds(GetCurrentTimestamp(), 30000);
/*
* Issue cancel request. Unfortunately, there's no good way to limit the
* amount of time that we might block inside PQgetCancel().
*/
if ((cancel = PQgetCancel(conn)))
{
if (!PQcancel(cancel, errbuf, sizeof(errbuf)))
{
ereport(WARNING,
(errcode(ERRCODE_CONNECTION_FAILURE),
errmsg("could not send cancel request: %s",
errbuf)));
PQfreeCancel(cancel);
return false;
}
PQfreeCancel(cancel);
}
/* Get and discard the result of the query. */
if (pgfdw_get_cleanup_result(conn, endtime, &result))
return false;
PQclear(result);
return true;
}
/*
* Submit a query during (sub)abort cleanup and wait up to 30 seconds for the
* result. If the query is executed without error, the return value is true.
* If the query is executed successfully but returns an error, the return
* value is true if and only if ignore_errors is set. If the query can't be
* sent or times out, the return value is false.
*
* It's not a huge problem if we throw an ERROR here, but if we get into error
* recursion trouble, we'll end up slamming the connection shut, which will
* necessitate failing the entire toplevel transaction even if subtransactions
* were used. Try to use WARNING where we can.
*/
static bool
pgfdw_exec_cleanup_query(PGconn *conn, const char *query, bool ignore_errors)
{
PGresult *result = NULL;
TimestampTz endtime;
/*
* If it takes too long to execute a cleanup query, assume the connection
* is dead. It's fairly likely that this is why we aborted in the first
* place (e.g. statement timeout, user cancel), so the timeout shouldn't
* be too long.
*/
endtime = TimestampTzPlusMilliseconds(GetCurrentTimestamp(), 30000);
/*
* Submit a query. Since we don't use non-blocking mode, this also can
* block. But its risk is relatively small, so we ignore that for now.
*/
if (!PQsendQuery(conn, query))
{
pgfdw_report_error(WARNING, NULL, conn, false, query);
return false;
}
/* Get the result of the query. */
if (pgfdw_get_cleanup_result(conn, endtime, &result))
return false;
/* Issue a warning if not successful. */
if (PQresultStatus(result) != PGRES_COMMAND_OK)
{
pgfdw_report_error(WARNING, result, conn, true, query);
return ignore_errors;
}
PQclear(result);
return true;
}
/*
* Get, during abort cleanup, the result of a query that is in progress. This
* might be a query that is being interrupted by transaction abort, or it might
* be a query that was initiated as part of transaction abort to get the remote
* side back to the appropriate state.
*
* endtime is the time at which we should give up and assume the remote
* side is dead. Returns true if the timeout expired, otherwise false.
* Sets *result except in case of a timeout.
*/
static bool
pgfdw_get_cleanup_result(PGconn *conn, TimestampTz endtime, PGresult **result)
{
volatile bool timed_out = false;
PGresult *volatile last_res = NULL;
/* In what follows, do not leak any PGresults on an error. */
PG_TRY();
{
for (;;)
{
PGresult *res;
while (PQisBusy(conn))
{
int wc;
TimestampTz now = GetCurrentTimestamp();
long cur_timeout;
/* If timeout has expired, give up, else get sleep time. */
cur_timeout = TimestampDifferenceMilliseconds(now, endtime);
if (cur_timeout <= 0)
{
timed_out = true;
goto exit;
}
/* Sleep until there's something to do */
wc = WaitLatchOrSocket(MyLatch,
WL_LATCH_SET | WL_SOCKET_READABLE | WL_TIMEOUT,
PQsocket(conn),
cur_timeout, PG_WAIT_EXTENSION);
ResetLatch(MyLatch);
CHECK_FOR_INTERRUPTS();
/* Data available in socket? */
if (wc & WL_SOCKET_READABLE)
{
if (!PQconsumeInput(conn))
{
/* connection trouble; treat the same as a timeout */
timed_out = true;
goto exit;
}
}
}
res = PQgetResult(conn);
if (res == NULL)
break; /* query is complete */
PQclear(last_res);
last_res = res;
}
exit: ;
}
PG_CATCH();
{
PQclear(last_res);
PG_RE_THROW();
}
PG_END_TRY();
if (timed_out)
PQclear(last_res);
else
*result = last_res;
return timed_out;
}
|