public inbox for [email protected]help / color / mirror / Atom feed
Re: POC: Carefully exposing information without authentication 5+ messages / 3 participants [nested] [flat]
* Re: POC: Carefully exposing information without authentication @ 2026-01-09 13:56 Antonin Houska <[email protected]> 2026-01-09 14:15 ` Re: POC: Carefully exposing information without authentication Greg Sabino Mullane <[email protected]> 0 siblings, 1 reply; 5+ messages in thread From: Antonin Houska @ 2026-01-09 13:56 UTC (permalink / raw) To: Greg Sabino Mullane <[email protected]>; +Cc: Tom Lane <[email protected]>; [email protected] <[email protected]> Greg Sabino Mullane <[email protected]> wrote: > Version 4 attached, rebased to account for new tests, plus a new instra-test > check to make sure LWP::UserAgent is available before running. I'm still not sure it's necessary to handle the problem at socket level. I imagine it can be implemented this way: 1. Add a new field to the PGconn structure, indicating that the client is only requesting the server status information, and adjust pg_isready so it sets this option. 2. Adjust libpq frontend (pqBuildStartupPacket3) so it adds the corresponding option to the startup packet. 3. On server, if ProcessStartupPacket() sees that option, call ereport(FATAL) with a specific error code, and let the appropriate GUCs control the contents of the error message. pg_isready would then just print out the message. I haven't tried to write any code, so it's possible that I'm missing something. Regarding configuration, I'd prefer a single GUC. The value can be a comma-separated list of keywords, each representing particular piece of information to be exposed. -- Antonin Houska Web: https://www.cybertec-postgresql.com ^ permalink raw reply [nested|flat] 5+ messages in thread
* Re: POC: Carefully exposing information without authentication 2026-01-09 13:56 Re: POC: Carefully exposing information without authentication Antonin Houska <[email protected]> @ 2026-01-09 14:15 ` Greg Sabino Mullane <[email protected]> 2026-02-17 19:42 ` Re: POC: Carefully exposing information without authentication Greg Sabino Mullane <[email protected]> 0 siblings, 1 reply; 5+ messages in thread From: Greg Sabino Mullane @ 2026-01-09 14:15 UTC (permalink / raw) To: Antonin Houska <[email protected]>; +Cc: Tom Lane <[email protected]>; [email protected] <[email protected]> On Fri, Jan 9, 2026 at 8:56 AM Antonin Houska <[email protected]> wrote: > 1. Add a new field to the PGconn structure This kind of defeats one of the major strengths of this patch, which is allowing systems that don't speak the protocol to get at this information. > Regarding configuration, I'd prefer a single GUC. The value can be a > comma-separated list of keywords, each representing particular piece of > information to be exposed. > Yes, I could see some advantages to that, although I still like the simplicity of separate boolean values. I've no strong feelings either way. Let's see if others weigh in. Thanks for looking over this patch! Cheers, Greg ^ permalink raw reply [nested|flat] 5+ messages in thread
* Re: POC: Carefully exposing information without authentication 2026-01-09 13:56 Re: POC: Carefully exposing information without authentication Antonin Houska <[email protected]> 2026-01-09 14:15 ` Re: POC: Carefully exposing information without authentication Greg Sabino Mullane <[email protected]> @ 2026-02-17 19:42 ` Greg Sabino Mullane <[email protected]> 2026-02-17 19:58 ` Re: POC: Carefully exposing information without authentication Andres Freund <[email protected]> 0 siblings, 1 reply; 5+ messages in thread From: Greg Sabino Mullane @ 2026-02-17 19:42 UTC (permalink / raw) To: Antonin Houska <[email protected]>; +Cc: Tom Lane <[email protected]>; [email protected] <[email protected]> Please find attached a rebased and lightly reworked version of this patch. The most significant change is the test file now uses IO::Socket::INET via $node->raw_connect. Also changed to allow case-insensitive calls, moved to a better docs group, moved the defines and typedefs up, and changed the exit to just a simple _exit() Cheers, Greg Attachments: [application/octet-stream] 0005-Allow-specific-information-to-be-output-directly-by-Postgres.patch (13.1K, 3-0005-Allow-specific-information-to-be-output-directly-by-Postgres.patch) download | inline diff: From 438d935387a9b58c85c2d566ca18a1ca60910832 Mon Sep 17 00:00:00 2001 From: Greg Sabino Mullane <[email protected]> Date: Tue, 17 Feb 2026 14:35:32 -0500 Subject: [PATCH] Allow specific information to be output directly by Postgres. --- src/backend/tcop/backend_startup.c | 185 ++++++++++++++++++ src/backend/utils/misc/guc_parameters.dat | 19 ++ src/backend/utils/misc/postgresql.conf.sample | 6 + src/include/postmaster/postmaster.h | 3 + src/test/modules/test_misc/meson.build | 1 + src/test/modules/test_misc/t/011_expose.pl | 122 ++++++++++++ 6 files changed, 336 insertions(+) create mode 100644 src/test/modules/test_misc/t/011_expose.pl diff --git a/src/backend/tcop/backend_startup.c b/src/backend/tcop/backend_startup.c index c517115927c..e7696d34d59 100644 --- a/src/backend/tcop/backend_startup.c +++ b/src/backend/tcop/backend_startup.c @@ -46,6 +46,29 @@ bool Trace_connection_negotiation = false; uint32 log_connections = 0; char *log_connections_string = NULL; +bool expose_recovery = false; +bool expose_sysid = false; +bool expose_version = false; + +#define EXPOSE_MIN_QUERY 9 /* Shortest possible line: "Get /info" */ +#define EXPOSE_MAX_QUERY 16 /* Longest possible GET line */ + +typedef enum +{ + EXPOSE_NOTHING, + EXPOSE_HEAD_REPLICA, + EXPOSE_GET_ALL, + EXPOSE_GET_REPLICA, + EXPOSE_GET_SYSID, + EXPOSE_GET_VERSION, +} ExposeReturnType; + +typedef struct +{ + const char *endpoint; + const bool *require; + ExposeReturnType type; +} endpoint_action; /* Other globals */ @@ -65,6 +88,7 @@ static void SendNegotiateProtocolVersion(List *unrecognized_protocol_options); static void process_startup_packet_die(SIGNAL_ARGS); static void StartupPacketTimeoutHandler(void); static bool validate_log_connections_options(List *elemlist, uint32 *flags); +static bool ExposeInformation(pgsocket fd); /* * Entry point for a new backend process. @@ -148,6 +172,14 @@ BackendInitialize(ClientSocket *client_sock, CAC_state cac) StringInfoData ps_data; MemoryContext oldcontext; + /* + * Scan for a simple GET / HEAD request. If this is detected and + * handled, we are done and can immediately exit + */ + if ((expose_recovery || expose_sysid || expose_version) + && ExposeInformation(client_sock->sock)) + _exit(0); /* Safe to use exit: no state or resources created yet */ + /* Tell fd.c about the long-lived FD associated with the client_sock */ ReserveExternalFD(); @@ -1125,3 +1157,156 @@ assign_log_connections(const char *newval, void *extra) { log_connections = *((int *) extra); } + +/* + * ExposeInformation + * + * Handle early socket probe before full backend startup. + * Responds to small set of predefined endpoints (e.g. GET /info) + * + * Requires at least one "expose_" GUC to be true. + * + * Returns true if any endpoint is recognized. + */ + +static bool +ExposeInformation(pgsocket fd) +{ + static endpoint_action endpoint_actions[] = + { + { + "HEAD /replica", &expose_recovery, EXPOSE_HEAD_REPLICA + }, + { + "GET /replica", &expose_recovery, EXPOSE_GET_REPLICA + }, + { + "GET /sysid", &expose_sysid, EXPOSE_GET_SYSID + }, + { + "GET /version", &expose_version, EXPOSE_GET_VERSION + }, + { + "GET /info", NULL, EXPOSE_GET_ALL + } + }; + + ssize_t n; + char buf[EXPOSE_MAX_QUERY + 1]; + ExposeReturnType type; + + Assert(expose_recovery || expose_sysid || expose_version); + + do + { + n = recv(fd, buf, EXPOSE_MAX_QUERY, MSG_PEEK); + } while (n < 0 && errno == EINTR); + + /* + * Leave as soon as possible if no chance we are interested. + * (we also leave on partial reads from slow clients) + * We also simply return false for n == -1 + */ + if (n < EXPOSE_MIN_QUERY) + return false; + + buf[n] = '\0'; + + type = EXPOSE_NOTHING; + for (int i = 0; i < lengthof(endpoint_actions); i++) + { + if ( + pg_strncasecmp(buf, endpoint_actions[i].endpoint, strlen(endpoint_actions[i].endpoint)) == 0 + && + (endpoint_actions[i].require == NULL + || + *(endpoint_actions[i].require) + )) + { + type = endpoint_actions[i].type; + break; + } + } + + if (type == EXPOSE_NOTHING) + return false; + + { + static const char http_version[] = "HTTP/1.1"; + static const char http_type[] = "Content-Type: text/plain"; + static const char http_conn[] = "Connection: close"; + static const char http_len[] = "Content-Length"; + + StringInfoData msg; + + if (type == EXPOSE_HEAD_REPLICA) + { + /* + * Caller only cares about the HTTP response code, so no content + * needed + */ + + initStringInfoExt(&msg, 64); + + appendStringInfo(&msg, + "%s %s\r\n" + "%s\r\n" + "%s\r\n\r\n", + http_version, + (RecoveryInProgress() ? "200 OK" : "503 Service Unavailable"), + http_type, + http_conn + ); + } + else + { + StringInfoData content; + + initStringInfoExt(&content, 64); + + if (expose_recovery && (type == EXPOSE_GET_ALL || type == EXPOSE_GET_REPLICA)) + appendStringInfo(&content, "%s%d\r\n", + type == EXPOSE_GET_ALL ? "RECOVERY: " : "", + RecoveryInProgress() ? 1 : 0); + if (expose_sysid && (type == EXPOSE_GET_ALL || type == EXPOSE_GET_SYSID)) + appendStringInfo(&content, "%s" UINT64_FORMAT "\r\n", + type == EXPOSE_GET_ALL ? "SYSID: " : "", + GetSystemIdentifier()); + if (expose_version && (type == EXPOSE_GET_ALL || type == EXPOSE_GET_VERSION)) + appendStringInfo(&content, "%s%d\r\n", + type == EXPOSE_GET_ALL ? "VERSION: " : "", + PG_VERSION_NUM); + + initStringInfoExt(&msg, 256); + + appendStringInfo(&msg, + "%s 200 OK\r\n" + "%s\r\n" + "%s: %d\r\n" + "%s\r\n\r\n" + "%s", + http_version, + http_type, + http_len, content.len, + http_conn, + content.data + ); + + pfree(content.data); + } + + do + { + n = send(fd, msg.data, msg.len, 0); + } while (n < 0 && errno == EINTR); + + pfree(msg.data); + + if (n < 0) + elog(DEBUG1, "could not send to client: %m"); + + return true; + + } + +} diff --git a/src/backend/utils/misc/guc_parameters.dat b/src/backend/utils/misc/guc_parameters.dat index 271c033952e..3e99d9f6b7c 100644 --- a/src/backend/utils/misc/guc_parameters.dat +++ b/src/backend/utils/misc/guc_parameters.dat @@ -1010,6 +1010,25 @@ boot_val => 'false', }, +{ name => 'expose_recovery', type => 'bool', context => 'PGC_SIGHUP', group => 'CONN_AUTH_AUTH', + short_desc => 'Exposes if the server is in recovery mode without a login.', + variable => 'expose_recovery', + boot_val => 'false', +}, + +{ name => 'expose_sysid', type => 'bool', context => 'PGC_SIGHUP', group => 'CONN_AUTH_AUTH', + short_desc => 'Exposes the system identifier without a login.', + variable => 'expose_sysid', + boot_val => 'false', +}, + +{ name => 'expose_version', type => 'bool', context => 'PGC_SIGHUP', group => 'CONN_AUTH_AUTH', + short_desc => 'Exposes the server version without a login.', + variable => 'expose_version', + boot_val => 'false', +}, + + { name => 'extension_control_path', type => 'string', context => 'PGC_SUSET', group => 'CLIENT_CONN_OTHER', short_desc => 'Sets the path for extension control files.', long_desc => 'The remaining extension script and secondary control files are then loaded from the same directory where the primary control file was found.', diff --git a/src/backend/utils/misc/postgresql.conf.sample b/src/backend/utils/misc/postgresql.conf.sample index f938cc65a3a..76b640e4878 100644 --- a/src/backend/utils/misc/postgresql.conf.sample +++ b/src/backend/utils/misc/postgresql.conf.sample @@ -91,6 +91,12 @@ # disconnection while running queries; # 0 for never +# - Expose information - + +#expose_recovery = off +#expose_sysid = off +#expose_version = off + # - Authentication - #authentication_timeout = 1min # 1s-600s diff --git a/src/include/postmaster/postmaster.h b/src/include/postmaster/postmaster.h index d6ab9ee2d96..b042336728f 100644 --- a/src/include/postmaster/postmaster.h +++ b/src/include/postmaster/postmaster.h @@ -70,6 +70,9 @@ extern PGDLLIMPORT bool restart_after_crash; extern PGDLLIMPORT bool remove_temp_files_after_crash; extern PGDLLIMPORT bool send_abort_for_crash; extern PGDLLIMPORT bool send_abort_for_kill; +extern PGDLLIMPORT bool expose_recovery; +extern PGDLLIMPORT bool expose_sysid; +extern PGDLLIMPORT bool expose_version; #ifdef WIN32 extern PGDLLIMPORT HANDLE PostmasterHandle; diff --git a/src/test/modules/test_misc/meson.build b/src/test/modules/test_misc/meson.build index 6e8db1621a7..c40a0455708 100644 --- a/src/test/modules/test_misc/meson.build +++ b/src/test/modules/test_misc/meson.build @@ -19,6 +19,7 @@ tests += { 't/008_replslot_single_user.pl', 't/009_log_temp_files.pl', 't/010_index_concurrently_upsert.pl', + 't/011_expose.pl', ], # The injection points are cluster-wide, so disable installcheck 'runningcheck': false, diff --git a/src/test/modules/test_misc/t/011_expose.pl b/src/test/modules/test_misc/t/011_expose.pl new file mode 100644 index 00000000000..3496e3ae283 --- /dev/null +++ b/src/test/modules/test_misc/t/011_expose.pl @@ -0,0 +1,122 @@ +# Copyright (c) 2026, PostgreSQL Global Development Group + +# Test gathering information before authentication via expose_* variables + +# Force use of TCP/IP - call before the 'use' +INIT{ $PostgreSQL::Test::Utils::use_unix_sockets = 0; } + +use strict; +use warnings; +use PostgreSQL::Test::Cluster; +use PostgreSQL::Test::Utils; +use Test::More; + +my $node = PostgreSQL::Test::Cluster->new('node1'); + +# Set as logical here so we can restart it as a replica later +$node->init(allows_streaming => 'logical'); +$node->start; + +my $server_version = $node->safe_psql('postgres', 'show server_version_num'); +my $bindir = $node->config_data('--bindir'); +my $datadir = $node->data_dir; +my $cdata = qx{$bindir/pg_controldata -D $datadir 2>&1}; +my ($sysid) = $cdata =~ /Database system identifier:\s+(\d+)/; +my $receive_length = 200; + +my ($socket, $response, $test); + +sub call_socket { + my $string = shift; + $socket->close() if defined $socket; + $socket = $node->raw_connect(); + $socket->send($string); + $response = ''; + select(undef, undef, undef, 0.1); + $socket->recv($response, $receive_length); + return; +} + +$test = 'GET /info returns nothing when nothing is listening'; +call_socket('GET /info'); +is ($response, '', $test); + +$test = 'HEAD /replica returns nothing when nothing is listening'; +call_socket('HEAD /replica'); +is ($response, '', $test); + +$node->append_conf('postgresql.conf', 'expose_recovery=on'); +$node->reload(); + +$test = 'GET /replica returns HTTP code 200 when expose_recovery is true (primary)'; +call_socket('GET /replica'); +like ($response, qr{^HTTP/1.1 200 }, $test); + +$test = 'GET /replica returns "0" when expose_recovery is true (primary)'; +like ($response, qr{\r\n0\r\n}, $test); + +$test = 'HEAD /replica returns HTTP code 503 when expose_recovery is true (primary)'; +call_socket('HEAD /replica'); +like ($response, qr{^HTTP/1.1 503 }, $test); + +$test = 'GET /info returns "RECOVERY: 0" when expose_recovery is true (primary)'; +call_socket('GET /info'); +like ($response, qr{RECOVERY: 0\r\n}, $test); + +$test = 'GET /info does not return version information when expose_version is false'; +unlike ($response, qr{VERSION}, $test); + +$test = 'GET /info does not return sysid information when expose_sysid is false'; +unlike ($response, qr{SYSID}, $test); + +$node->append_conf('postgresql.conf', 'expose_version=on'); +$node->append_conf('postgresql.conf', 'expose_sysid=on'); +$node->reload(); + +$test = 'GET /info returns correct version when expose_version is true'; +call_socket('GET /info'); +like ($response, qr/VERSION: $server_version/, $test); + +$test = 'GET /info returns correct value when expose_sysid is true'; +like ($response, qr/SYSID: $sysid/, $test); + +$test = 'Get /sysid returns correct value when expose_sysid is true'; +call_socket('Get /sysid'); ## Not required to be all uppercase according to the spec! +like ($response, qr/^$sysid\r\n/m, $test); + +$test = 'GET /version returns correct value when expose_version is true'; +call_socket('GET /version'); +like ($response, qr/^$server_version\r\n/m, $test); + +$test = 'GET /foobar returns nothing'; +call_socket('GET /foobar'); +is ($response, '', $test); + +$node->set_standby_mode(); +$node->restart(); + +$test = 'GET /replica returns HTTP code 200 when expose_recovery is true (replica)'; +call_socket('GET /replica'); +like ($response, qr{^HTTP/1.1 200 }, $test); + +$test = 'GET /replica returns "1" when expose_recovery is true (replica)'; +like ($response, qr{^1\r\n}m, $test); + +$test = 'HEAD /replica returns HTTP code 200 when expose_recovery is true (replica)'; +call_socket('HEAD /replica'); +like ($response, qr{^HTTP/1.1 200 }, $test); + +$test = 'GET /info returns "RECOVERY: 1" when expose_recovery is true (replica)'; +call_socket('GET /info'); +like ($response, qr/RECOVERY: 1/, $test); + +$node->append_conf('postgresql.conf', 'expose_version=off'); +$node->reload(); + +$test = 'GET /version returns nothing after expose_version turned back off'; +call_socket('GET /version'); +is ($response, '', $test); + +$socket->close(); + +done_testing(); -- 2.47.3 ^ permalink raw reply [nested|flat] 5+ messages in thread
* Re: POC: Carefully exposing information without authentication 2026-01-09 13:56 Re: POC: Carefully exposing information without authentication Antonin Houska <[email protected]> 2026-01-09 14:15 ` Re: POC: Carefully exposing information without authentication Greg Sabino Mullane <[email protected]> 2026-02-17 19:42 ` Re: POC: Carefully exposing information without authentication Greg Sabino Mullane <[email protected]> @ 2026-02-17 19:58 ` Andres Freund <[email protected]> 2026-03-25 01:59 ` Re: POC: Carefully exposing information without authentication Greg Sabino Mullane <[email protected]> 0 siblings, 1 reply; 5+ messages in thread From: Andres Freund @ 2026-02-17 19:58 UTC (permalink / raw) To: Greg Sabino Mullane <[email protected]>; +Cc: Antonin Houska <[email protected]>; Tom Lane <[email protected]>; [email protected] <[email protected]> Hi, On 2026-02-17 14:42:48 -0500, Greg Sabino Mullane wrote: > Subject: [PATCH] Allow specific information to be output directly by Postgres. I strongly encourage you to include a justification for why this is desirable, so a casual reviewer doesn't have to reread the thread. > @@ -148,6 +172,14 @@ BackendInitialize(ClientSocket *client_sock, CAC_state cac) > StringInfoData ps_data; > MemoryContext oldcontext; > > + /* > + * Scan for a simple GET / HEAD request. If this is detected and > + * handled, we are done and can immediately exit > + */ > + if ((expose_recovery || expose_sysid || expose_version) > + && ExposeInformation(client_sock->sock)) > + _exit(0); /* Safe to use exit: no state or resources created yet */ > + > /* Tell fd.c about the long-lived FD associated with the client_sock */ > ReserveExternalFD(); > What about direct TLS connections? How can a cluster coordinator trust unauthenticated plain text communication that can just be man-in-the-middled? It's not obvious that it's a good idea to expose this on the same socket as normal client connections. IMO you'd want to limit this to a smaller set of interfaces than normal client connections. > +/* > + * ExposeInformation > + * > + * Handle early socket probe before full backend startup. > + * Responds to small set of predefined endpoints (e.g. GET /info) > + * > + * Requires at least one "expose_" GUC to be true. > + * > + * Returns true if any endpoint is recognized. > + */ > + > +static bool > +ExposeInformation(pgsocket fd) > +{ > + static endpoint_action endpoint_actions[] = > + { > + { > + "HEAD /replica", &expose_recovery, EXPOSE_HEAD_REPLICA > + }, > + { > + "GET /replica", &expose_recovery, EXPOSE_GET_REPLICA > + }, > + { > + "GET /sysid", &expose_sysid, EXPOSE_GET_SYSID > + }, > + { > + "GET /version", &expose_version, EXPOSE_GET_VERSION > + }, > + { > + "GET /info", NULL, EXPOSE_GET_ALL > + } > + }; > + > + ssize_t n; > + char buf[EXPOSE_MAX_QUERY + 1]; > + ExposeReturnType type; > + > + Assert(expose_recovery || expose_sysid || expose_version); > + > + do > + { > + n = recv(fd, buf, EXPOSE_MAX_QUERY, MSG_PEEK); > + } while (n < 0 && errno == EINTR); > > + /* > + * Leave as soon as possible if no chance we are interested. > + * (we also leave on partial reads from slow clients) > + * We also simply return false for n == -1 > + */ > + if (n < EXPOSE_MIN_QUERY) > + return false; IIRC the socket is in blocking mode at this point (that's only changed in pq_init()), therefore this might actually block? While it's unlikely, I don't see any guarantee that a single receive would actually get the whole message from the client either, so this seems like it might fail spuriously. > diff --git a/src/backend/utils/misc/guc_parameters.dat b/src/backend/utils/misc/guc_parameters.dat > index 271c033952e..3e99d9f6b7c 100644 > --- a/src/backend/utils/misc/guc_parameters.dat > +++ b/src/backend/utils/misc/guc_parameters.dat > @@ -1010,6 +1010,25 @@ > boot_val => 'false', > }, > > +{ name => 'expose_recovery', type => 'bool', context => 'PGC_SIGHUP', group => 'CONN_AUTH_AUTH', > + short_desc => 'Exposes if the server is in recovery mode without a login.', > + variable => 'expose_recovery', > + boot_val => 'false', > +}, > + > +{ name => 'expose_sysid', type => 'bool', context => 'PGC_SIGHUP', group => 'CONN_AUTH_AUTH', > + short_desc => 'Exposes the system identifier without a login.', > + variable => 'expose_sysid', > + boot_val => 'false', > +}, > + > +{ name => 'expose_version', type => 'bool', context => 'PGC_SIGHUP', group => 'CONN_AUTH_AUTH', > + short_desc => 'Exposes the server version without a login.', > + variable => 'expose_version', > + boot_val => 'false', > +}, > + If we were to do this, I'd recommend a single expose GUC that has the different values as a comma separated list, instead a growing list of GUCs. Greetings, Andres Freund ^ permalink raw reply [nested|flat] 5+ messages in thread
* Re: POC: Carefully exposing information without authentication 2026-01-09 13:56 Re: POC: Carefully exposing information without authentication Antonin Houska <[email protected]> 2026-01-09 14:15 ` Re: POC: Carefully exposing information without authentication Greg Sabino Mullane <[email protected]> 2026-02-17 19:42 ` Re: POC: Carefully exposing information without authentication Greg Sabino Mullane <[email protected]> 2026-02-17 19:58 ` Re: POC: Carefully exposing information without authentication Andres Freund <[email protected]> @ 2026-03-25 01:59 ` Greg Sabino Mullane <[email protected]> 0 siblings, 0 replies; 5+ messages in thread From: Greg Sabino Mullane @ 2026-03-25 01:59 UTC (permalink / raw) To: Andres Freund <[email protected]>; +Cc: Antonin Houska <[email protected]>; Tom Lane <[email protected]>; [email protected] <[email protected]> Thank you for looking over this. New version attached. On Tue, Feb 17, 2026 at 2:58 PM Andres Freund <[email protected]> wrote: > What about direct TLS connections? Not handled. How can a cluster coordinator trust unauthenticated plain text > communication that can just be man-in-the-middled? > They cannot. But that's why this is only exposing non-critical information. Right now the security scanners that are banging on port 5432 and scraping the returned error lines are not worried about man-in-the-middle. :) Obviously, if your threat model is people capturing and modifying non-encrypted traffic to your Postgres server, you would not use this. It's not obvious that it's a good idea to expose this on the same socket as > normal client connections. IMO you'd want to limit this to a smaller set > of interfaces than normal client connections. > I'm not entirely clear what that smaller set would mean in practice. > IIRC the socket is in blocking mode at this point (that's only changed in > pq_init()), therefore this might actually block? While it's unlikely, I > don't see any guarantee that a single receive would actually get the whole > message from the client either, so this seems like it might fail spuriously. > Yes, there are some very unlikely edge cases, but this is meant to be good enough, not a perfectly bulletproof HTTP server. Clients should try again on failures. Which if they do occur for this trivial amount of traffic probably indicates much bigger problems. If we were to do this, I'd recommend a single expose GUC that has the > different values as a comma separated list, instead a growing list of GUCs. > Done - see attached for a new version which consolidates the bools into a single comma-separated GUC called "expose_information". I also added some docs, and changed the "replica" to return "REPLICA" instead of "RECOVERY". I like the latter better, but replica lines up better with existing tools. -- Cheers, Greg Attachments: [application/octet-stream] 0006-Allow-specific-information-to-be-output-directly-by-Postgres.patch (19.1K, 3-0006-Allow-specific-information-to-be-output-directly-by-Postgres.patch) download | inline diff: From 257aaa706cb272148d877d0f67fce2bc1e49a39b Mon Sep 17 00:00:00 2001 From: Greg Sabino Mullane <[email protected]> Date: Tue, 24 Mar 2026 21:15:23 -0400 Subject: [PATCH] Allow-specific-information-to-be-output-directly-by-Postgres --- doc/src/sgml/config.sgml | 61 ++++ src/backend/tcop/backend_startup.c | 262 ++++++++++++++++++ src/backend/utils/misc/guc_parameters.dat | 10 + src/backend/utils/misc/postgresql.conf.sample | 7 + src/include/tcop/backend_startup.h | 2 + src/include/utils/guc_hooks.h | 2 + src/test/modules/test_misc/meson.build | 1 + src/test/modules/test_misc/t/011_expose.pl | 121 ++++++++ 8 files changed, 466 insertions(+) create mode 100644 src/test/modules/test_misc/t/011_expose.pl diff --git a/doc/src/sgml/config.sgml b/doc/src/sgml/config.sgml index 8cdd826fbd3..a32c7e7f6a0 100644 --- a/doc/src/sgml/config.sgml +++ b/doc/src/sgml/config.sgml @@ -643,6 +643,67 @@ include_dir 'conf.d' <variablelist> + <varlistentry id="guc-expose-information" xreflabel="expose_information"> + <term><varname>expose_information</varname> (<type>string</type>) + <indexterm> + <primary><varname>expose_information</varname> configuration parameter</primary> + </indexterm> + </term> + <listitem> + <para> + Allows for specific information to be returned from the servers without + requiring a login. Requests should come in as a simple HTTP request as a + GET or HEAD to the PostgreSQL port. + The default is the empty string, <literal>''</literal>, which + prevents any information from being output. The following options may be + specified alone or in a comma-separated list: + </para> + + <table id="expose-information-options"> + <title>Expose Information Options</title> + <tgroup cols="2"> + <colspec colname="col1" colwidth="1*"/> + <colspec colname="col2" colwidth="2*"/> + <thead> + <row> + <entry>Name</entry> + <entry>Description</entry> + </row> + </thead> + <tbody> + <row> + <entry><literal>replica</literal></entry> + <entry>Reports if the server is a replica (i.e. is in recovery mode) or not. If the request is <literal>HEAD /replica</literal>, + then an HTTP response code of 200 (yes it is a replica) or 503 (not a replica) is returned. This + can be used as a drop-in replacement for the same functionality provided by the Patroni program. + For the request <literal>GET /replica</literal> or <literal>GET /info</literal>, + the string <literal>REPLICA: 1</literal> or <literal>REPLICA: 0</literal> is returned.</entry> + </row> + + <row> + <entry><literal>sysid</literal></entry> + <entry>Returns the system identifier of the server. This can be useful to determine if the underlying + server has changed, as the initdb program will always generate a new system identifier. + For the request <literal>GET /sysid</literal> or <literal>GET /info</literal>, + the string <literal>SYSID: 12345</literal> is returned, in which "12345" will be + the specific system identifier (typically a 20-digit number)</entry> + </row> + + <row> + <entry><literal>version</literal></entry> + <entry>Returns the current version of the server. Specifically, the value of + <literal>server_version_num</literal>. + For the request <literal>GET /version</literal> or <literal>GET /info</literal>, + the string <literal>VERSION: 190000</literal> is returned (for this example, + the version of Postgres is 19.0)</entry> + </row> + </tbody> + </tgroup> + </table> + + </listitem> + </varlistentry> + <varlistentry id="guc-listen-addresses" xreflabel="listen_addresses"> <term><varname>listen_addresses</varname> (<type>string</type>) <indexterm> diff --git a/src/backend/tcop/backend_startup.c b/src/backend/tcop/backend_startup.c index 5abf276c898..271ea9ddcb8 100644 --- a/src/backend/tcop/backend_startup.c +++ b/src/backend/tcop/backend_startup.c @@ -46,6 +46,33 @@ bool Trace_connection_negotiation = false; uint32 log_connections = 0; char *log_connections_string = NULL; +int Expose_information = 0; +char *Expose_information_string = NULL; + +/* Expose information bitmap */ +#define EXPOSE_INFO_REPLICA 1 +#define EXPOSE_INFO_SYSID 2 +#define EXPOSE_INFO_VERSION 4 + +#define EXPOSE_MIN_QUERY 9 /* Shortest possible line: "Get /info" */ +#define EXPOSE_MAX_QUERY 16 /* Longest possible GET line */ + +typedef enum +{ + EXPOSE_NOTHING, + EXPOSE_HEAD_REPLICA, + EXPOSE_GET_ALL, + EXPOSE_GET_REPLICA, + EXPOSE_GET_SYSID, + EXPOSE_GET_VERSION, +} ExposeReturnType; + +typedef struct +{ + const char *endpoint; + int require; + ExposeReturnType type; +} endpoint_action; /* Other globals */ @@ -65,6 +92,7 @@ static void SendNegotiateProtocolVersion(List *unrecognized_protocol_options); static void process_startup_packet_die(SIGNAL_ARGS); static void StartupPacketTimeoutHandler(void); static bool validate_log_connections_options(List *elemlist, uint32 *flags); +static bool ExposeInformation(pgsocket fd); /* * Entry point for a new backend process. @@ -148,6 +176,15 @@ BackendInitialize(ClientSocket *client_sock, CAC_state cac) StringInfoData ps_data; MemoryContext oldcontext; + /* + * Scan for a simple GET / HEAD request. If this is detected and handled, + * we are done and can immediately exit. + */ + if ((Expose_information > 0) + && ExposeInformation(client_sock->sock)) + _exit(0); /* Safe to use exit: no state or resources + * created yet */ + /* Tell fd.c about the long-lived FD associated with the client_sock */ ReserveExternalFD(); @@ -1075,6 +1112,72 @@ next: ; } +/* + * GUC check_hook for expose_information + */ +bool +check_expose_information(char **newval, void **extra, GucSource source) +{ + char *rawstring; + List *elemlist; + ListCell *l; + int newexpose = 0; + int *myextra; + + /* Need a modifiable copy of string */ + rawstring = pstrdup(*newval); + + /* Parse string into list of identifiers */ + if (!SplitIdentifierString(rawstring, ',', &elemlist)) + { + /* syntax error in list */ + GUC_check_errdetail("List syntax is invalid."); + pfree(rawstring); + list_free(elemlist); + return false; + } + + foreach(l, elemlist) + { + char *tok = (char *) lfirst(l); + + if (pg_strcasecmp(tok, "replica") == 0) + newexpose |= EXPOSE_INFO_REPLICA; + else if (pg_strcasecmp(tok, "sysid") == 0) + newexpose |= EXPOSE_INFO_SYSID; + else if (pg_strcasecmp(tok, "version") == 0) + newexpose |= EXPOSE_INFO_VERSION; + else + { + GUC_check_errdetail("Unrecognized key word: \"%s\".", tok); + pfree(rawstring); + list_free(elemlist); + return false; + } + } + + pfree(rawstring); + list_free(elemlist); + + myextra = (int *) guc_malloc(LOG, sizeof(int)); + if (!myextra) + return false; + *myextra = newexpose; + *extra = myextra; + + return true; +} + +/* + * GUC assign_hook for expose_information + */ +void +assign_expose_information(const char *newval, void *extra) +{ + Expose_information = *((int *) extra); +} + + /* * GUC check hook for log_connections */ @@ -1127,3 +1230,162 @@ assign_log_connections(const char *newval, void *extra) { log_connections = *((int *) extra); } + +/* + * ExposeInformation + * + * Handle early socket probe before full backend startup. + * Responds to small set of predefined endpoints (e.g. GET /info) + * + * Requires the expose_information GUC to be non-empty + * + * Returns true if any endpoint is recognized. + */ + +static bool +ExposeInformation(pgsocket fd) +{ + static const endpoint_action endpoint_actions[] = + { + { + "HEAD /replica", EXPOSE_INFO_REPLICA, EXPOSE_HEAD_REPLICA + }, + { + "GET /replica", EXPOSE_INFO_REPLICA, EXPOSE_GET_REPLICA + }, + { + "GET /sysid", EXPOSE_INFO_SYSID, EXPOSE_GET_SYSID + }, + { + "GET /version", EXPOSE_INFO_VERSION, EXPOSE_GET_VERSION + }, + { + "GET /info", 0, EXPOSE_GET_ALL + } + }; + + ssize_t n; + char buf[EXPOSE_MAX_QUERY + 1]; + ExposeReturnType type; + + Assert(Expose_information > 0); + + do + { + n = recv(fd, buf, EXPOSE_MAX_QUERY, MSG_PEEK); + } while (n < 0 && errno == EINTR); + + /* + * Leave as soon as possible if no chance we are interested. We also leave + * on partial reads from slow clients. Note that we return false for n == + * -1 + */ + if (n < EXPOSE_MIN_QUERY) + return false; + + buf[n] = '\0'; + + type = EXPOSE_NOTHING; + for (int i = 0; i < lengthof(endpoint_actions); i++) + { + if ( + pg_strncasecmp(buf, endpoint_actions[i].endpoint, strlen(endpoint_actions[i].endpoint)) == 0 + && + ((endpoint_actions[i].require == 0) + || + (Expose_information & endpoint_actions[i].require) + )) + { + type = endpoint_actions[i].type; + break; + } + } + + if (type == EXPOSE_NOTHING) + return false; + + { + static const char http_version[] = "HTTP/1.1"; + static const char http_type[] = "Content-Type: text/plain"; + static const char http_conn[] = "Connection: close"; + static const char http_len[] = "Content-Length"; + + StringInfoData msg; + + if (type == EXPOSE_HEAD_REPLICA) + { + /* + * Caller only cares about the HTTP response code, so no content + * needed + */ + + initStringInfoExt(&msg, 64); + + appendStringInfo(&msg, + "%s %s\r\n" + "%s\r\n" + "%s\r\n\r\n", + http_version, + (RecoveryInProgress() ? "200 OK" : "503 Service Unavailable"), + http_type, + http_conn + ); + } + else + { + StringInfoData content; + + initStringInfoExt(&content, 64); + + if ((Expose_information & EXPOSE_INFO_REPLICA) + && + (type == EXPOSE_GET_ALL || type == EXPOSE_GET_REPLICA)) + appendStringInfo(&content, "%s%d\r\n", + type == EXPOSE_GET_ALL ? "REPLICA: " : "", + RecoveryInProgress() ? 1 : 0); + if ((Expose_information & EXPOSE_INFO_SYSID) + && + (type == EXPOSE_GET_ALL || type == EXPOSE_GET_SYSID)) + appendStringInfo(&content, "%s" UINT64_FORMAT "\r\n", + type == EXPOSE_GET_ALL ? "SYSID: " : "", + GetSystemIdentifier()); + if ((Expose_information & EXPOSE_INFO_VERSION) + && + (type == EXPOSE_GET_ALL || type == EXPOSE_GET_VERSION)) + appendStringInfo(&content, "%s%d\r\n", + type == EXPOSE_GET_ALL ? "VERSION: " : "", + PG_VERSION_NUM); + + initStringInfoExt(&msg, 256); + + appendStringInfo(&msg, + "%s 200 OK\r\n" + "%s\r\n" + "%s: %d\r\n" + "%s\r\n\r\n" + "%s", + http_version, + http_type, + http_len, content.len, + http_conn, + content.data + ); + + pfree(content.data); + } + + do + { + n = send(fd, msg.data, msg.len, 0); + } while (n < 0 && errno == EINTR); + + pfree(msg.data); + + if (n < 0) + elog(DEBUG1, "could not send to client: %m"); + + return true; + + } + +} diff --git a/src/backend/utils/misc/guc_parameters.dat b/src/backend/utils/misc/guc_parameters.dat index 0c9854ad8fc..8c8ad0d8d0a 100644 --- a/src/backend/utils/misc/guc_parameters.dat +++ b/src/backend/utils/misc/guc_parameters.dat @@ -1010,6 +1010,16 @@ boot_val => 'false', }, +{ name => 'expose_information', type => 'string', context => 'PGC_SIGHUP', group => 'CONN_AUTH_AUTH', + short_desc => 'Expose limited information without needing to login', + long_desc => 'Valid values are combinations of "replica", "sysid", and "version"', + flags => 'GUC_LIST_INPUT', + variable => 'Expose_information_string', + boot_val => '""', + check_hook => 'check_expose_information', + assign_hook => 'assign_expose_information', +}, + { name => 'extension_control_path', type => 'string', context => 'PGC_SUSET', group => 'CLIENT_CONN_OTHER', short_desc => 'Sets the path for extension control files.', long_desc => 'The remaining extension script and secondary control files are then loaded from the same directory where the primary control file was found.', diff --git a/src/backend/utils/misc/postgresql.conf.sample b/src/backend/utils/misc/postgresql.conf.sample index e4abe6c0077..5e361fd4b58 100644 --- a/src/backend/utils/misc/postgresql.conf.sample +++ b/src/backend/utils/misc/postgresql.conf.sample @@ -93,6 +93,13 @@ # disconnection while running queries; # 0 for never +# - Expose information - + +expose_information = '' # comma-separated list of items to expose + # replica = if the server is in recovery or not + # sysid = the current system identifier for this server + # version = the current version of this server + # - Authentication - #authentication_timeout = 1min # 1s-600s diff --git a/src/include/tcop/backend_startup.h b/src/include/tcop/backend_startup.h index d486f926319..6204dd98e81 100644 --- a/src/include/tcop/backend_startup.h +++ b/src/include/tcop/backend_startup.h @@ -20,6 +20,8 @@ extern PGDLLIMPORT bool Trace_connection_negotiation; extern PGDLLIMPORT uint32 log_connections; extern PGDLLIMPORT char *log_connections_string; +extern PGDLLIMPORT int Expose_information; +extern PGDLLIMPORT char *Expose_information_string; /* Other globals */ extern PGDLLIMPORT struct ConnectionTiming conn_timing; diff --git a/src/include/utils/guc_hooks.h b/src/include/utils/guc_hooks.h index b01697c1f60..04d9f024d26 100644 --- a/src/include/utils/guc_hooks.h +++ b/src/include/utils/guc_hooks.h @@ -62,6 +62,8 @@ extern void assign_default_text_search_config(const char *newval, void *extra); extern bool check_default_with_oids(bool *newval, void **extra, GucSource source); extern const char *show_effective_wal_level(void); +extern bool check_expose_information(char **newval, void **extra, GucSource source); +extern void assign_expose_information(const char *newval, void *extra); extern bool check_huge_page_size(int *newval, void **extra, GucSource source); extern void assign_io_method(int newval, void *extra); extern bool check_io_max_concurrency(int *newval, void **extra, GucSource source); diff --git a/src/test/modules/test_misc/meson.build b/src/test/modules/test_misc/meson.build index 6e8db1621a7..c40a0455708 100644 --- a/src/test/modules/test_misc/meson.build +++ b/src/test/modules/test_misc/meson.build @@ -19,6 +19,7 @@ tests += { 't/008_replslot_single_user.pl', 't/009_log_temp_files.pl', 't/010_index_concurrently_upsert.pl', + 't/011_expose.pl', ], # The injection points are cluster-wide, so disable installcheck 'runningcheck': false, diff --git a/src/test/modules/test_misc/t/011_expose.pl b/src/test/modules/test_misc/t/011_expose.pl new file mode 100644 index 00000000000..df97b98096b --- /dev/null +++ b/src/test/modules/test_misc/t/011_expose.pl @@ -0,0 +1,121 @@ +# Copyright (c) 2026, PostgreSQL Global Development Group + +# Test gathering information before authentication via expose_* variables + +# Force use of TCP/IP - must be called before the 'use' section +INIT{ $PostgreSQL::Test::Utils::use_unix_sockets = 0; } + +use strict; +use warnings; +use PostgreSQL::Test::Cluster; +use PostgreSQL::Test::Utils; +use Test::More; + +my $node = PostgreSQL::Test::Cluster->new('node1'); + +# Set as logical here so we can restart it as a replica later +$node->init(allows_streaming => 'logical'); +$node->start; + +my $server_version = $node->safe_psql('postgres', 'show server_version_num'); +my $bindir = $node->config_data('--bindir'); +my $datadir = $node->data_dir; +my $cdata = qx{$bindir/pg_controldata -D $datadir 2>&1}; +my ($sysid) = $cdata =~ /Database system identifier:\s+(\d+)/; +my $receive_length = 200; + +my ($socket, $response, $test); + +sub call_socket { + my $string = shift; + $socket->close() if defined $socket; + $socket = $node->raw_connect(); + $socket->send($string); + $response = ''; + select(undef, undef, undef, 0.1); + $socket->recv($response, $receive_length); + return; +} + +$test = 'GET /info returns nothing when nothing is listening'; +call_socket('GET /info'); +is ($response, '', $test); + +$test = 'HEAD /replica returns nothing when nothing is listening'; +call_socket('HEAD /replica'); +is ($response, '', $test); + +$node->append_conf('postgresql.conf', "expose_information = 'replica'"); +$node->reload(); + +$test = q{GET /replica returns HTTP code 200 when expose_information contains 'replica' (primary)}; +call_socket('GET /replica'); +like ($response, qr{^HTTP/1.1 200 }, $test); + +$test = q{GET /replica returns "0" when expose_information contains 'replica' (primary)}; +like ($response, qr{\r\n0\r\n}, $test); + +$test = q{HEAD /replica returns HTTP code 503 when expose_information contains 'replica' (primary)}; +call_socket('HEAD /replica'); +like ($response, qr{^HTTP/1.1 503 }, $test); + +$test = q{GET /info returns "REPLICA: 0" when expose_information contains 'replica' (primary)}; +call_socket('GET /info'); +like ($response, qr{REPLICA: 0\r\n}, $test); + +$test = q{GET /info does not return version information when expose_information does not contain 'version'}; +unlike ($response, qr{VERSION}, $test); + +$test = q{GET /info does not return sysid information when expose_information does not contain 'sysid'}; +unlike ($response, qr{SYSID}, $test); + +$node->append_conf('postgresql.conf', "expose_information= 'replica,sysid,version'"); +$node->reload(); + +$test = q{GET /info returns correct version when expose_information contains 'version'}; +call_socket('GET /info'); +like ($response, qr/VERSION: $server_version/, $test); + +$test = q{GET /info returns correct value when expose_information contains 'sysid'}; +like ($response, qr/SYSID: $sysid/, $test); + +$test = q{Get /sysid returns correct value when expose_information contains 'sysid'}; +call_socket('Get /sysid'); ## Not required to be all uppercase according to the spec! +like ($response, qr/^$sysid\r\n/m, $test); + +$test = q{GET /version returns correct value when expose_information contains 'version'}; +call_socket('GET /version'); +like ($response, qr/^$server_version\r\n/m, $test); + +$test = 'GET /foobar returns nothing'; +call_socket('GET /foobar'); +is ($response, '', $test); + +$node->set_standby_mode(); +$node->restart(); + +$test = q{GET /replica returns HTTP code 200 when expose_information contains 'replica' (replica)}; +call_socket('GET /replica'); +like ($response, qr{^HTTP/1.1 200 }, $test); + +$test = q{GET /replica returns "1" when expose_information contains 'replica' (replica)}; +like ($response, qr{^1\r\n}m, $test); + +$test = q{HEAD /replica returns HTTP code 200 when expose_information contains 'replica' (replica)}; +call_socket('HEAD /replica'); +like ($response, qr{^HTTP/1.1 200 }, $test); + +$test = q{GET /info returns "REPLICA: 1" when expose_information contains 'replica' (replica)}; +call_socket('GET /info'); +like ($response, qr/REPLICA: 1/, $test); + +$node->append_conf('postgresql.conf', "expose_information=''"); +$node->reload(); + +$test = q{GET /version returns nothing after expose_information no longer has 'version'}; +call_socket('GET /version'); +is ($response, '', $test); + +$socket->close(); + +done_testing(); -- 2.47.3 ^ permalink raw reply [nested|flat] 5+ messages in thread
end of thread, other threads:[~2026-03-25 01:59 UTC | newest] Thread overview: 5+ messages (download: mbox mbox.gz follow: Atom feed) -- links below jump to the message on this page -- 2026-01-09 13:56 Re: POC: Carefully exposing information without authentication Antonin Houska <[email protected]> 2026-01-09 14:15 ` Greg Sabino Mullane <[email protected]> 2026-02-17 19:42 ` Greg Sabino Mullane <[email protected]> 2026-02-17 19:58 ` Andres Freund <[email protected]> 2026-03-25 01:59 ` Greg Sabino Mullane <[email protected]>
This inbox is served by agora; see mirroring instructions for how to clone and mirror all data and code used for this inbox