Hello,
Attached patch implements the following TODO item :
Track number of WAL files ready to be archived in pg_stat_archiver
However, it will track the total number of any file ready to be
archived, not only WAL files.
Please let me know what you think about it.
Regards.
--
Julien Rouhaud
http://dalibo.com - http://dalibo.org
*** a/doc/src/sgml/monitoring.sgml
--- b/doc/src/sgml/monitoring.sgml
***************
*** 728,733 **** postgres 27093 0.0 0.0 30096 2752 ? Ss 11:34 0:00 postgres: ser
--- 728,738 ----
<entry>Time of the last failed archival operation</entry>
</row>
<row>
+ <entry><structfield>ready_count</></entry>
+ <entry><type>bigint</type></entry>
+ <entry>Number of files waiting to be archived</entry>
+ </row>
+ <row>
<entry><structfield>stats_reset</></entry>
<entry><type>timestamp with time zone</type></entry>
<entry>Time at which these statistics were last reset</entry>
*** a/src/backend/access/transam/xlogarchive.c
--- b/src/backend/access/transam/xlogarchive.c
***************
*** 24,29 ****
--- 24,30 ----
#include "access/xlog_internal.h"
#include "miscadmin.h"
#include "postmaster/startup.h"
+ #include "pgstat.h"
#include "replication/walsender.h"
#include "storage/fd.h"
#include "storage/ipc.h"
***************
*** 539,544 **** XLogArchiveNotify(const char *xlog)
--- 540,548 ----
/* Notify archiver that it's got something to do */
if (IsUnderPostmaster)
SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+
+ /* Tell the collector about a new file waiting to be archived */
+ pgstat_send_archiver(xlog, ARCH_READY);
}
/*
*** a/src/backend/catalog/system_views.sql
--- b/src/backend/catalog/system_views.sql
***************
*** 697,702 **** CREATE VIEW pg_stat_archiver AS
--- 697,703 ----
s.failed_count,
s.last_failed_wal,
s.last_failed_time,
+ s.ready_count,
s.stats_reset
FROM pg_stat_get_archiver() s;
*** a/src/backend/postmaster/pgarch.c
--- b/src/backend/postmaster/pgarch.c
***************
*** 491,497 **** pgarch_ArchiverCopyLoop(void)
* Tell the collector about the WAL file that we successfully
* archived
*/
! pgstat_send_archiver(xlog, false);
break; /* out of inner retry loop */
}
--- 491,497 ----
* Tell the collector about the WAL file that we successfully
* archived
*/
! pgstat_send_archiver(xlog, ARCH_SUCCESS);
break; /* out of inner retry loop */
}
***************
*** 501,507 **** pgarch_ArchiverCopyLoop(void)
* Tell the collector about the WAL file that we failed to
* archive
*/
! pgstat_send_archiver(xlog, true);
if (++failures >= NUM_ARCHIVE_RETRIES)
{
--- 501,507 ----
* Tell the collector about the WAL file that we failed to
* archive
*/
! pgstat_send_archiver(xlog, ARCH_FAIL);
if (++failures >= NUM_ARCHIVE_RETRIES)
{
*** a/src/backend/postmaster/pgstat.c
--- b/src/backend/postmaster/pgstat.c
***************
*** 36,41 ****
--- 36,42 ----
#include "access/transam.h"
#include "access/twophase_rmgr.h"
#include "access/xact.h"
+ #include "access/xlog_internal.h"
#include "catalog/pg_database.h"
#include "catalog/pg_proc.h"
#include "lib/ilist.h"
***************
*** 3084,3094 **** pgstat_send(void *msg, int len)
* pgstat_send_archiver() -
*
* Tell the collector about the WAL file that we successfully
! * archived or failed to archive.
* ----------
*/
void
! pgstat_send_archiver(const char *xlog, bool failed)
{
PgStat_MsgArchiver msg;
--- 3085,3096 ----
* pgstat_send_archiver() -
*
* Tell the collector about the WAL file that we successfully
! * archived or failed to archive, or the new file waiting
! * to be archived.
* ----------
*/
void
! pgstat_send_archiver(const char *xlog, ArchiverReason reason)
{
PgStat_MsgArchiver msg;
***************
*** 3096,3102 **** pgstat_send_archiver(const char *xlog, bool failed)
* Prepare and send the message
*/
pgstat_setheader(&msg.m_hdr, PGSTAT_MTYPE_ARCHIVER);
! msg.m_failed = failed;
strncpy(msg.m_xlog, xlog, sizeof(msg.m_xlog));
msg.m_timestamp = GetCurrentTimestamp();
pgstat_send(&msg, sizeof(msg));
--- 3098,3104 ----
* Prepare and send the message
*/
pgstat_setheader(&msg.m_hdr, PGSTAT_MTYPE_ARCHIVER);
! msg.m_reason = reason;
strncpy(msg.m_xlog, xlog, sizeof(msg.m_xlog));
msg.m_timestamp = GetCurrentTimestamp();
pgstat_send(&msg, sizeof(msg));
***************
*** 3921,3927 **** pgstat_read_statsfiles(Oid onlydb, bool permanent, bool deep)
/*
* Try to open the stats file. If it doesn't exist, the backends simply
* return zero for anything and the collector simply starts from scratch
! * with empty counters.
*
* ENOENT is a possibility if the stats collector is not running or has
* not yet written the stats file the first time. Any other failure
--- 3923,3930 ----
/*
* Try to open the stats file. If it doesn't exist, the backends simply
* return zero for anything and the collector simply starts from scratch
! * with empty counters, except for the .ready files count which should
! * always give the real number of files.
*
* ENOENT is a possibility if the stats collector is not running or has
* not yet written the stats file the first time. Any other failure
***************
*** 3934,3939 **** pgstat_read_statsfiles(Oid onlydb, bool permanent, bool deep)
--- 3937,3970 ----
(errcode_for_file_access(),
errmsg("could not open statistics file \"%s\": %m",
statfile)));
+
+ /* Initialize the archive ready counter */
+ char XLogArchiveStatusDir[MAXPGPATH];
+ DIR *rldir;
+ struct dirent *rlde;
+
+ snprintf(XLogArchiveStatusDir, MAXPGPATH, XLOGDIR "/archive_status");
+ rldir = AllocateDir(XLogArchiveStatusDir);
+ if (rldir == NULL)
+ ereport(ERROR,
+ (errcode_for_file_access(),
+ errmsg("could not open archive status directory \"%s\": %m",
+ XLogArchiveStatusDir)));
+
+ while ((rlde = ReadDir(rldir, XLogArchiveStatusDir)) != NULL)
+ {
+ int basenamelen = (int) strlen(rlde->d_name) - 6;
+
+ if (basenamelen >= MIN_XFN_CHARS &&
+ basenamelen <= MAX_XFN_CHARS &&
+ strspn(rlde->d_name, VALID_XFN_CHARS) >= basenamelen &&
+ strcmp(rlde->d_name + basenamelen, ".ready") == 0)
+ {
+ ++archiverStats.ready_count;
+ }
+ }
+ FreeDir(rldir);
+
return dbhash;
}
***************
*** 4842,4849 **** pgstat_recv_resetsharedcounter(PgStat_MsgResetsharedcounter *msg, int len)
}
else if (msg->m_resettarget == RESET_ARCHIVER)
{
! /* Reset the archiver statistics for the cluster. */
memset(&archiverStats, 0, sizeof(archiverStats));
archiverStats.stat_reset_timestamp = GetCurrentTimestamp();
}
--- 4873,4887 ----
}
else if (msg->m_resettarget == RESET_ARCHIVER)
{
! PgStat_Counter ready_count;
! /*
! * Reset the archiver statistics for the cluster.
! * We must keep the ready_count value as it should
! * always reflect the real count.
! */
! ready_count = archiverStats.ready_count;
memset(&archiverStats, 0, sizeof(archiverStats));
+ archiverStats.ready_count = ready_count;
archiverStats.stat_reset_timestamp = GetCurrentTimestamp();
}
***************
*** 4984,5004 **** pgstat_recv_analyze(PgStat_MsgAnalyze *msg, int len)
static void
pgstat_recv_archiver(PgStat_MsgArchiver *msg, int len)
{
! if (msg->m_failed)
! {
! /* Failed archival attempt */
! ++archiverStats.failed_count;
! memcpy(archiverStats.last_failed_wal, msg->m_xlog,
! sizeof(archiverStats.last_failed_wal));
! archiverStats.last_failed_timestamp = msg->m_timestamp;
! }
! else
{
! /* Successful archival operation */
! ++archiverStats.archived_count;
! memcpy(archiverStats.last_archived_wal, msg->m_xlog,
! sizeof(archiverStats.last_archived_wal));
! archiverStats.last_archived_timestamp = msg->m_timestamp;
}
}
--- 5022,5048 ----
static void
pgstat_recv_archiver(PgStat_MsgArchiver *msg, int len)
{
! switch (msg->m_reason)
{
! case ARCH_FAIL:
! /* Failed archival attempt */
! ++archiverStats.failed_count;
! memcpy(archiverStats.last_failed_wal, msg->m_xlog,
! sizeof(archiverStats.last_failed_wal));
! archiverStats.last_failed_timestamp = msg->m_timestamp;
! break;
! case ARCH_SUCCESS:
! /* Successful archival operation */
! ++archiverStats.archived_count;
! memcpy(archiverStats.last_archived_wal, msg->m_xlog,
! sizeof(archiverStats.last_archived_wal));
! archiverStats.last_archived_timestamp = msg->m_timestamp;
! --archiverStats.ready_count;
! break;
! case ARCH_READY:
! /* New file waiting to be archived */
! ++archiverStats.ready_count;
! break;
}
}
*** a/src/backend/utils/adt/pgstatfuncs.c
--- b/src/backend/utils/adt/pgstatfuncs.c
***************
*** 1746,1752 **** pg_stat_get_archiver(PG_FUNCTION_ARGS)
MemSet(nulls, 0, sizeof(nulls));
/* Initialise attributes information in the tuple descriptor */
! tupdesc = CreateTemplateTupleDesc(7, false);
TupleDescInitEntry(tupdesc, (AttrNumber) 1, "archived_count",
INT8OID, -1, 0);
TupleDescInitEntry(tupdesc, (AttrNumber) 2, "last_archived_wal",
--- 1746,1752 ----
MemSet(nulls, 0, sizeof(nulls));
/* Initialise attributes information in the tuple descriptor */
! tupdesc = CreateTemplateTupleDesc(8, false);
TupleDescInitEntry(tupdesc, (AttrNumber) 1, "archived_count",
INT8OID, -1, 0);
TupleDescInitEntry(tupdesc, (AttrNumber) 2, "last_archived_wal",
***************
*** 1759,1765 **** pg_stat_get_archiver(PG_FUNCTION_ARGS)
TEXTOID, -1, 0);
TupleDescInitEntry(tupdesc, (AttrNumber) 6, "last_failed_time",
TIMESTAMPTZOID, -1, 0);
! TupleDescInitEntry(tupdesc, (AttrNumber) 7, "stats_reset",
TIMESTAMPTZOID, -1, 0);
BlessTupleDesc(tupdesc);
--- 1759,1767 ----
TEXTOID, -1, 0);
TupleDescInitEntry(tupdesc, (AttrNumber) 6, "last_failed_time",
TIMESTAMPTZOID, -1, 0);
! TupleDescInitEntry(tupdesc, (AttrNumber) 7, "ready_count",
! INT8OID, -1, 0);
! TupleDescInitEntry(tupdesc, (AttrNumber) 8, "stats_reset",
TIMESTAMPTZOID, -1, 0);
BlessTupleDesc(tupdesc);
***************
*** 1790,1799 **** pg_stat_get_archiver(PG_FUNCTION_ARGS)
else
values[5] = TimestampTzGetDatum(archiver_stats->last_failed_timestamp);
if (archiver_stats->stat_reset_timestamp == 0)
! nulls[6] = true;
else
! values[6] = TimestampTzGetDatum(archiver_stats->stat_reset_timestamp);
/* Returns the record as Datum */
PG_RETURN_DATUM(HeapTupleGetDatum(
--- 1792,1802 ----
else
values[5] = TimestampTzGetDatum(archiver_stats->last_failed_timestamp);
+ values[6] = Int64GetDatum(archiver_stats->ready_count);
if (archiver_stats->stat_reset_timestamp == 0)
! nulls[7] = true;
else
! values[7] = TimestampTzGetDatum(archiver_stats->stat_reset_timestamp);
/* Returns the record as Datum */
PG_RETURN_DATUM(HeapTupleGetDatum(
*** a/src/include/catalog/pg_proc.h
--- b/src/include/catalog/pg_proc.h
***************
*** 2751,2757 **** DATA(insert OID = 2844 ( pg_stat_get_db_blk_read_time PGNSP PGUID 12 1 0 0 0 f
DESCR("statistics: block read time, in msec");
DATA(insert OID = 2845 ( pg_stat_get_db_blk_write_time PGNSP PGUID 12 1 0 0 0 f f f f t f s 1 0 701 "26" _null_ _null_ _null_ _null_ pg_stat_get_db_blk_write_time _null_ _null_ _null_ ));
DESCR("statistics: block write time, in msec");
! DATA(insert OID = 3195 ( pg_stat_get_archiver PGNSP PGUID 12 1 0 0 0 f f f f f f s 0 0 2249 "" "{20,25,1184,20,25,1184,1184}" "{o,o,o,o,o,o,o}" "{archived_count,last_archived_wal,last_archived_time,failed_count,last_failed_wal,last_failed_time,stats_reset}" _null_ pg_stat_get_archiver _null_ _null_ _null_ ));
DESCR("statistics: information about WAL archiver");
DATA(insert OID = 2769 ( pg_stat_get_bgwriter_timed_checkpoints PGNSP PGUID 12 1 0 0 0 f f f f t f s 0 0 20 "" _null_ _null_ _null_ _null_ pg_stat_get_bgwriter_timed_checkpoints _null_ _null_ _null_ ));
DESCR("statistics: number of timed checkpoints started by the bgwriter");
--- 2751,2757 ----
DESCR("statistics: block read time, in msec");
DATA(insert OID = 2845 ( pg_stat_get_db_blk_write_time PGNSP PGUID 12 1 0 0 0 f f f f t f s 1 0 701 "26" _null_ _null_ _null_ _null_ pg_stat_get_db_blk_write_time _null_ _null_ _null_ ));
DESCR("statistics: block write time, in msec");
! DATA(insert OID = 3195 ( pg_stat_get_archiver PGNSP PGUID 12 1 0 0 0 f f f f f f s 0 0 2249 "" "{20,25,1184,20,25,1184,20,1184}" "{o,o,o,o,o,o,o,o}" "{archived_count,last_archived_wal,last_archived_time,failed_count,last_failed_wal,last_failed_time,ready_count,stats_reset}" _null_ pg_stat_get_archiver _null_ _null_ _null_ ));
DESCR("statistics: information about WAL archiver");
DATA(insert OID = 2769 ( pg_stat_get_bgwriter_timed_checkpoints PGNSP PGUID 12 1 0 0 0 f f f f t f s 0 0 20 "" _null_ _null_ _null_ _null_ pg_stat_get_bgwriter_timed_checkpoints _null_ _null_ _null_ ));
DESCR("statistics: number of timed checkpoints started by the bgwriter");
*** a/src/include/pgstat.h
--- b/src/include/pgstat.h
***************
*** 376,382 **** typedef struct PgStat_MsgAnalyze
typedef struct PgStat_MsgArchiver
{
PgStat_MsgHdr m_hdr;
! bool m_failed; /* Failed attempt */
char m_xlog[MAX_XFN_CHARS + 1];
TimestampTz m_timestamp;
} PgStat_MsgArchiver;
--- 376,382 ----
typedef struct PgStat_MsgArchiver
{
PgStat_MsgHdr m_hdr;
! int m_reason;
char m_xlog[MAX_XFN_CHARS + 1];
TimestampTz m_timestamp;
} PgStat_MsgArchiver;
***************
*** 651,656 **** typedef struct PgStat_ArchiverStats
--- 651,657 ----
char last_failed_wal[MAX_XFN_CHARS + 1]; /* WAL file involved in
* last failure */
TimestampTz last_failed_timestamp; /* last archival failure time */
+ PgStat_Counter ready_count; /* Number of files waiting to be archived */
TimestampTz stat_reset_timestamp;
} PgStat_ArchiverStats;
***************
*** 690,695 **** typedef enum BackendState
--- 691,707 ----
} BackendState;
/* ----------
+ * Archiver reason
+ * ----------
+ */
+ typedef enum ArchiverReason
+ {
+ ARCH_SUCCESS,
+ ARCH_FAIL,
+ ARCH_READY,
+ } ArchiverReason;
+
+ /* ----------
* Shared-memory data structures
* ----------
*/
***************
*** 934,940 **** extern void pgstat_twophase_postcommit(TransactionId xid, uint16 info,
extern void pgstat_twophase_postabort(TransactionId xid, uint16 info,
void *recdata, uint32 len);
! extern void pgstat_send_archiver(const char *xlog, bool failed);
extern void pgstat_send_bgwriter(void);
/* ----------
--- 946,952 ----
extern void pgstat_twophase_postabort(TransactionId xid, uint16 info,
void *recdata, uint32 len);
! extern void pgstat_send_archiver(const char *xlog, ArchiverReason reason);
extern void pgstat_send_bgwriter(void);
/* ----------
*** a/src/test/regress/expected/rules.out
--- b/src/test/regress/expected/rules.out
***************
*** 1659,1666 **** pg_stat_archiver| SELECT s.archived_count,
s.failed_count,
s.last_failed_wal,
s.last_failed_time,
s.stats_reset
! FROM pg_stat_get_archiver() s(archived_count, last_archived_wal, last_archived_time, failed_count, last_failed_wal, last_failed_time, stats_reset);
pg_stat_bgwriter| SELECT pg_stat_get_bgwriter_timed_checkpoints() AS checkpoints_timed,
pg_stat_get_bgwriter_requested_checkpoints() AS checkpoints_req,
pg_stat_get_checkpoint_write_time() AS checkpoint_write_time,
--- 1659,1667 ----
s.failed_count,
s.last_failed_wal,
s.last_failed_time,
+ s.ready_count,
s.stats_reset
! FROM pg_stat_get_archiver() s(archived_count, last_archived_wal, last_archived_time, failed_count, last_failed_wal, last_failed_time, ready_count, stats_reset);
pg_stat_bgwriter| SELECT pg_stat_get_bgwriter_timed_checkpoints() AS checkpoints_timed,
pg_stat_get_bgwriter_requested_checkpoints() AS checkpoints_req,
pg_stat_get_checkpoint_write_time() AS checkpoint_write_time,
--
Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers
