ckdb - don't store sharesummaries in the DB and delay mark processing until sync has completed

src/ckdb.c

@@ -268,7 +268,7 @@ bool dbload_only_sharesummary = false;
  *  markersummaries and pplns payouts may not be correct */
 bool sharesummary_marks_limit = false;
 
-// DB users,workers load is complete
+// DB optioncontrol,users,workers,useratts load is complete
 bool db_users_complete = false;
 // DB load is complete
 bool db_load_complete = false;
@@ -758,7 +758,9 @@ static bool getdata1()
 		goto matane;
 	if (!(ok = users_fill(conn)))
 		goto matane;
-	ok = workers_fill(conn);
+	if (!(ok = workers_fill(conn)))
+		goto matane;
+	ok = useratts_fill(conn);
 
 matane:
 
@@ -796,21 +798,15 @@ static bool getdata3()
 	}
 	if (!(ok = workinfo_fill(conn)) || everyone_die)
 		goto sukamudai;
-	/* marks must be loaded before sharesummary
-	 * since sharesummary looks at the marks data */
 	if (!(ok = marks_fill(conn)) || everyone_die)
 		goto sukamudai;
+	/* must be after workinfo */
 	if (!(ok = workmarkers_fill(conn)) || everyone_die)
 		goto sukamudai;
 	if (!(ok = markersummary_fill(conn)) || everyone_die)
 		goto sukamudai;
-	if (!(ok = sharesummary_fill(conn)) || everyone_die)
-		goto sukamudai;
-	if (!confirm_sharesummary) {
-		if (!(ok = useratts_fill(conn)) || everyone_die)
-			goto sukamudai;
+	if (!confirm_sharesummary)
 		ok = poolstats_fill(conn);
-	}
 
 sukamudai:
 
@@ -828,32 +824,25 @@ static bool reload()
 	char *reason;
 	FILE *fp;
 
-	tv_to_buf(&(dbstatus.oldest_sharesummary_firstshare_n), buf, sizeof(buf));
-	LOGWARNING("%s(): %s oldest DB incomplete sharesummary", __func__, buf);
-	tv_to_buf(&(dbstatus.newest_sharesummary_firstshare_ay), buf, sizeof(buf));
-	LOGWARNING("%s(): %s newest DB complete sharesummary", __func__, buf);
+	tv_to_buf(&(dbstatus.newest_createdate_workmarker_workinfo),
+		  buf, sizeof(buf));
+	LOGWARNING("%s(): %s newest DB workmarker wid %"PRId64,
+		   __func__, buf,
+		   dbstatus.newest_workmarker_workinfoid);
 	tv_to_buf(&(dbstatus.newest_createdate_workinfo), buf, sizeof(buf));
-	LOGWARNING("%s(): %s newest DB workinfo", __func__, buf);
+	LOGWARNING("%s(): %s newest DB workinfo wid %"PRId64,
+		   __func__, buf, dbstatus.newest_workinfoid);
 	tv_to_buf(&(dbstatus.newest_createdate_poolstats), buf, sizeof(buf));
-	LOGWARNING("%s(): %s newest DB poolstats", __func__, buf);
+	LOGWARNING("%s(): %s newest DB poolstats (ignored)", __func__, buf);
 	tv_to_buf(&(dbstatus.newest_createdate_blocks), buf, sizeof(buf));
 	LOGWARNING("%s(): %s newest DB blocks (ignored)", __func__, buf);
 
-	if (dbstatus.oldest_sharesummary_firstshare_n.tv_sec)
-		copy_tv(&(dbstatus.sharesummary_firstshare), &(dbstatus.oldest_sharesummary_firstshare_n));
-	else
-		copy_tv(&(dbstatus.sharesummary_firstshare), &(dbstatus.newest_sharesummary_firstshare_ay));
-
-	copy_tv(&start, &(dbstatus.sharesummary_firstshare));
-	reason = "sharesummary";
+	copy_tv(&start, &(dbstatus.newest_createdate_workmarker_workinfo));
+	reason = "workmarkers";
 	if (!tv_newer(&start, &(dbstatus.newest_createdate_workinfo))) {
 		copy_tv(&start, &(dbstatus.newest_createdate_workinfo));
 		reason = "workinfo";
 	}
-	if (!tv_newer(&start, &(dbstatus.newest_createdate_poolstats))) {
-		copy_tv(&start, &(dbstatus.newest_createdate_poolstats));
-		reason = "poolstats";
-	}
 
 	tv_to_buf(&start, buf, sizeof(buf));
 	LOGWARNING("%s() restart timestamp %s for %s", __func__, buf, reason);
@@ -901,15 +890,18 @@ static bool write_pid(ckpool_t *ckp, const char *path, pid_t pid)
 		fclose(fp);
 		if (ret == 1 && !(kill(oldpid, 0))) {
 			if (!ckp->killold) {
-				LOGEMERG("Process %s pid %d still exists, start ckpool with -k if you wish to kill it",
+				LOGEMERG("Process %s pid %d still exists, start"
+					 " ckpool with -k if you wish to kill it",
 					 path, oldpid);
 				return false;
 			}
 			if (kill(oldpid, 9)) {
-				LOGEMERG("Unable to kill old process %s pid %d", path, oldpid);
+				LOGEMERG("Unable to kill old process %s pid %d",
+					 path, oldpid);
 				return false;
 			}
-			LOGWARNING("Killing off old process %s pid %d", path, oldpid);
+			LOGWARNING("Killing off old process %s pid %d",
+				   path, oldpid);
 		}
 	}
 	fp = fopen(path, "we");
@@ -3076,7 +3068,7 @@ static void *summariser(__maybe_unused void *arg)
 
 	rename_proc("db_summariser");
 
-	while (!everyone_die && !startup_complete)
+	while (!everyone_die && !reload_queue_complete)
 		cksleep_ms(42);
 
 	summariser_using_data = true;
@@ -3568,7 +3560,7 @@ static void *marker(__maybe_unused void *arg)
 
 	rename_proc("db_marker");
 
-	while (!everyone_die && !startup_complete)
+	while (!everyone_die && !reload_queue_complete)
 		cksleep_ms(42);
 
 	if (sharesummary_marks_limit) {
@@ -3579,16 +3571,6 @@ static void *marker(__maybe_unused void *arg)
 
 	marker_using_data = true;
 
-/* TODO: trigger this every workinfo change?
- *  note that history catch up would also mean the tigger would
- *  catch up at most 100 missing marks per shift
- *  however, also, a workinfo change means a sharesummary DB update,
- *  so would be best to (usually) wait until that is done
- * OR: avoid writing the sharesummaries to the DB at all
- *  and only write the markersummaries? - since 100 workinfoid shifts
- *  will usually mean that markersummaries are less than every hour
- *  (and a reload processes more than an hour) */
-
 	while (!everyone_die) {
 		for (i = 0; i < 5; i++) {
 			if (!everyone_die)
@@ -4439,6 +4421,7 @@ static bool reload_from(tv_t *start)
 	reloading = true;
 
 	copy_tv(&reload_timestamp, start);
+	// Go back further - one reload file
 	reload_timestamp.tv_sec -= reload_timestamp.tv_sec % ROLL_S;
 
 	tv_to_buf(start, buf, sizeof(buf));
@@ -4790,6 +4773,7 @@ static void *listener(void *arg)
 	return NULL;
 }
 
+#if 0
 /* TODO: This will be way faster traversing both trees simultaneously
  *  rather than traversing one and searching the other, then repeating
  *  in reverse. Will change it later */
@@ -4893,6 +4877,7 @@ static void compare_summaries(K_TREE *leftsum, char *leftname,
 			diff_first, diff_last, cd_buf1, cd_buf2);
 	}
 }
+#endif
 
 /* TODO: have a seperate option to find/store missing workinfo/shares/etc
  *  from the reload files, in a supplied UTC time range
@@ -4905,6 +4890,9 @@ static void compare_summaries(K_TREE *leftsum, char *leftname,
  *  and the payment is now wrong */
 static void confirm_reload()
 {
+#if 0
+	TODO: redo this using workmarkers
+
 	K_TREE *sharesummary_workinfoid_save;
 	__maybe_unused K_TREE *sharesummary_save;
 	__maybe_unused K_TREE *workinfo_save;
@@ -5219,6 +5207,7 @@ static void confirm_reload()
 	compare_summaries(sharesummary_workinfoid_root, "ReLoad",
 			  sharesummary_workinfoid_save, "DB",
 			  true, false);
+#endif
 }
 
 // TODO: handle workmarkers/markersummaries

src/ckdb.h

@@ -55,7 +55,7 @@
 
 #define DB_VLOCK "1"
 #define DB_VERSION "1.0.0"
-#define CKDB_VERSION DB_VERSION"-1.094"
+#define CKDB_VERSION DB_VERSION"-1.100"
 
 #define WHERE_FFL " - from %s %s() line %d"
 #define WHERE_FFL_HERE __FILE__, __func__, __LINE__
@@ -125,20 +125,12 @@ extern const char *addrpatt;
 #define MAX_PAYADDR '~'
 
 typedef struct loadstatus {
-	tv_t oldest_sharesummary_firstshare_n;
-	tv_t newest_sharesummary_firstshare_a;
-	tv_t newest_sharesummary_firstshare_ay;
-	tv_t sharesummary_firstshare; // whichever of above 2 used
-	tv_t oldest_sharesummary_firstshare_a;
-	tv_t newest_sharesummary_firstshare_y;
+	int64_t newest_workmarker_workinfoid;
+	int64_t newest_workinfoid;
+	tv_t newest_createdate_workmarker_workinfo;
 	tv_t newest_createdate_workinfo;
 	tv_t newest_createdate_poolstats;
-	tv_t newest_starttimeband_userstats;
 	tv_t newest_createdate_blocks;
-	int64_t oldest_workinfoid_n; // of oldest firstshare sharesummary n
-	int64_t oldest_workinfoid_a; // of oldest firstshare sharesummary a
-	int64_t newest_workinfoid_a; // of newest firstshare sharesummary a
-	int64_t newest_workinfoid_y; // of newest firstshare sharesummary y
 } LOADSTATUS;
 extern LOADSTATUS dbstatus;
 
@@ -1366,10 +1358,6 @@ typedef struct sharesummary {
 	double sharerej;
 	int64_t sharecount;
 	int64_t errorcount;
-	int64_t countlastupdate; // non-DB field
-	bool inserted; // non-DB field
-	bool saveaged; // non-DB field
-	bool reset; // non-DB field
 	tv_t firstshare;
 	tv_t lastshare;
 	double lastdiffacc;
@@ -2131,10 +2119,10 @@ extern cmp_t _cmp_height(char *coinbase1a, char *coinbase1b, WHERE_FFL_ARGS);
 extern cmp_t cmp_workinfo_height(K_ITEM *a, K_ITEM *b);
 extern K_ITEM *find_workinfo(int64_t workinfoid, K_TREE_CTX *ctx);
 extern K_ITEM *next_workinfo(int64_t workinfoid, K_TREE_CTX *ctx);
-extern bool workinfo_age(PGconn *conn, int64_t workinfoid, char *poolinstance,
-			 char *by, char *code, char *inet, tv_t *cd,
-			 tv_t *ss_first, tv_t *ss_last, int64_t *ss_count,
-			 int64_t *s_count, int64_t *s_diff);
+extern bool workinfo_age(int64_t workinfoid, char *poolinstance, char *by,
+			 char *code, char *inet, tv_t *cd, tv_t *ss_first,
+			 tv_t *ss_last, int64_t *ss_count, int64_t *s_count,
+			 int64_t *s_diff);
 extern cmp_t cmp_shares(K_ITEM *a, K_ITEM *b);
 extern cmp_t cmp_shareerrors(K_ITEM *a, K_ITEM *b);
 extern void dsp_sharesummary(K_ITEM *item, FILE *stream);
@@ -2152,8 +2140,8 @@ extern void zero_sharesummary(SHARESUMMARY *row, tv_t *cd, double diff);
 extern K_ITEM *_find_sharesummary(int64_t userid, char *workername,
 				  int64_t workinfoid, bool pool);
 extern K_ITEM *find_last_sharesummary(int64_t userid, char *workername);
-extern void auto_age_older(PGconn *conn, int64_t workinfoid, char *poolinstance,
-			   char *by, char *code, char *inet, tv_t *cd);
+extern void auto_age_older(int64_t workinfoid, char *poolinstance, char *by,
+			   char *code, char *inet, tv_t *cd);
 #define dbhash2btchash(_hash, _buf, _siz) \
 	_dbhash2btchash(_hash, _buf, _siz, WHERE_FFL_HERE)
 void _dbhash2btchash(char *hash, char *buf, size_t siz, WHERE_FFL_ARGS);
@@ -2245,6 +2233,8 @@ extern void _userinfo_block(BLOCKS *blocks, bool isnew, bool lock);
 			(_res) == PGRES_TUPLES_OK || \
 			(_res) == PGRES_EMPTY_QUERY)
 
+#define SQL_UNIQUE_VIOLATION "23505"
+
 #define CKPQ_READ true
 #define CKPQ_WRITE false
 
@@ -2348,11 +2338,12 @@ extern bool sharesummaries_to_markersummaries(PGconn *conn, WORKMARKERS *workmar
 						char *by, char *code, char *inet,
 						tv_t *cd, K_TREE *trf_root);
 extern char *ooo_status(char *buf, size_t siz);
-#define sharesummary_update(_conn, _s_row, _e_row, _ss_item, _by, _code, _inet, _cd) \
-		_sharesummary_update(_conn, _s_row, _e_row, _ss_item, _by, _code, _inet, _cd, \
+#define sharesummary_update(_s_row, _e_row, _ss_item, _by, _code, _inet, _cd) \
+		_sharesummary_update(_s_row, _e_row, _ss_item, _by, _code, _inet, _cd, \
 					WHERE_FFL_HERE)
-extern bool _sharesummary_update(PGconn *conn, SHARES *s_row, SHAREERRORS *e_row, K_ITEM *ss_item,
-				char *by, char *code, char *inet, tv_t *cd, WHERE_FFL_ARGS);
+extern bool _sharesummary_update(SHARES *s_row, SHAREERRORS *e_row, K_ITEM *ss_item,
+				 char *by, char *code, char *inet, tv_t *cd,
+				 WHERE_FFL_ARGS);
 extern bool sharesummary_fill(PGconn *conn);
 extern bool blocks_stats(PGconn *conn, int32_t height, char *blockhash,
 			 double diffacc, double diffinv, double shareacc,

src/ckdb_cmd.c

@@ -683,7 +683,8 @@ static char *cmd_poolstats_do(PGconn *conn, char *cmd, char *id, char *by,
 					by, code, inet, cd, igndup, trf_root);
 
 	if (!ok) {
-		LOGERR("%s() %s.failed.DBE", __func__, id);
+		if (!igndup)
+			LOGERR("%s() %s.failed.DBE", __func__, id);
 		return strdup("failed.DBE");
 	}
 	LOGDEBUG("%s.ok.", id);
@@ -698,13 +699,11 @@ static char *cmd_poolstats(PGconn *conn, char *cmd, char *id,
 {
 	bool igndup = false;
 
-	// confirm_summaries() doesn't call this
-	if (reloading) {
-		if (tv_equal(cd, &(dbstatus.newest_createdate_blocks)))
-			igndup = true;
-		else if (tv_newer(cd, &(dbstatus.newest_createdate_blocks)))
-			return NULL;
-	}
+	/* confirm_summaries() doesn't call this
+	 * We don't care about dups during reload since poolstats_fill()
+	 * doesn't load all the data */
+	if (reloading)
+		igndup = true;
 
 	return cmd_poolstats_do(conn, cmd, id, by, code, inet, cd, igndup, trf_root);
 }
@@ -1965,20 +1964,18 @@ static char *cmd_sharelog(PGconn *conn, char *cmd, char *id,
 		K_ITEM *i_ntime, *i_reward;
 		bool igndup = false;
 
-		if (reloading && !confirm_sharesummary) {
-			if (tv_equal(cd, &(dbstatus.newest_createdate_workinfo)))
-				igndup = true;
-			else if (tv_newer(cd, &(dbstatus.newest_createdate_workinfo)))
-				return NULL;
-		}
-
 		i_workinfoid = require_name(trf_root, "workinfoid", 1, NULL, reply, siz);
 		if (!i_workinfoid)
 			return strdup(reply);
 
-		if (confirm_sharesummary) {
-			TXT_TO_BIGINT("workinfoid", transfer_data(i_workinfoid), workinfoid);
+		TXT_TO_BIGINT("workinfoid", transfer_data(i_workinfoid), workinfoid);
 
+		if (reloading && !confirm_sharesummary) {
+			if (workinfoid <= dbstatus.newest_workinfoid)
+				igndup = true;
+		}
+
+		if (confirm_sharesummary) {
 			if (workinfoid < confirm_first_workinfoid ||
 			    workinfoid > confirm_last_workinfoid)
 				goto wiconf;
@@ -2056,12 +2053,6 @@ wiconf:
 		K_ITEM *i_secondaryuserid;
 		bool ok;
 
-		// This just excludes the shares we certainly don't need
-		if (reloading && !confirm_sharesummary) {
-			if (tv_newer(cd, &(dbstatus.sharesummary_firstshare)))
-				return NULL;
-		}
-
 		i_nonce = require_name(trf_root, "nonce", 1, NULL, reply, siz);
 		if (!i_nonce)
 			return strdup(reply);
@@ -2070,9 +2061,28 @@ wiconf:
 		if (!i_workinfoid)
 			return strdup(reply);
 
-		if (confirm_sharesummary) {
-			TXT_TO_BIGINT("workinfoid", transfer_data(i_workinfoid), workinfoid);
+		TXT_TO_BIGINT("workinfoid", transfer_data(i_workinfoid), workinfoid);
 
+		if (reloading && !confirm_sharesummary) {
+			/* ISDR (Ignored shares during reload)
+			 * This will discard any shares older than the newest
+			 *  workinfoidend of any workmarker - including ready
+			 *  but not processed workmarkers
+			 * This means that if a workmarker needs re-processing
+			 *  and all of it's shares need to be redone, that will
+			 *  require a seperate procedure to the reload
+			 *  This would be the (as yet non-existant)
+			 *   confirm_markersummary which will replace the
+			 *   now unusable confirm_sharesummary code
+			 * However, if the workmarker simply just needs to be
+			 *  flagged as processed, this avoids the problem of
+			 *  duplicating shares before flagging it
+			 */
+			if (workinfoid <= dbstatus.newest_workmarker_workinfoid)
+				return NULL;
+		}
+
+		if (confirm_sharesummary) {
 			if (workinfoid < confirm_first_workinfoid ||
 			    workinfoid > confirm_last_workinfoid)
 				goto sconf;
@@ -2151,12 +2161,6 @@ sconf:
 		K_ITEM *i_error, *i_secondaryuserid;
 		bool ok;
 
-		// This just excludes the shareerrors we certainly don't need
-		if (reloading && !confirm_sharesummary) {
-			if (tv_newer(cd, &(dbstatus.sharesummary_firstshare)))
-				return NULL;
-		}
-
 		i_username = require_name(trf_root, "username", 1, NULL, reply, siz);
 		if (!i_username)
 			return strdup(reply);
@@ -2165,6 +2169,13 @@ sconf:
 		if (!i_workinfoid)
 			return strdup(reply);
 
+		TXT_TO_BIGINT("workinfoid", transfer_data(i_workinfoid), workinfoid);
+
+		if (reloading && !confirm_sharesummary) {
+			// See comment 'ISDR' above for shares
+			if (workinfoid <= dbstatus.newest_workmarker_workinfoid)
+				return NULL;
+		}
 		if (confirm_sharesummary) {
 			TXT_TO_BIGINT("workinfoid", transfer_data(i_workinfoid), workinfoid);
 
@@ -2216,15 +2227,18 @@ seconf:
 		tv_t ss_first, ss_last;
 		bool ok;
 
-		if (reloading && !confirm_sharesummary) {
-			if (tv_newer(cd, &(dbstatus.sharesummary_firstshare)))
-				return NULL;
-		}
-
 		i_workinfoid = require_name(trf_root, "workinfoid", 1, NULL, reply, siz);
 		if (!i_workinfoid)
 			return strdup(reply);
 
+		TXT_TO_BIGINT("workinfoid", transfer_data(i_workinfoid), workinfoid);
+
+		if (reloading && !confirm_sharesummary) {
+			// This excludes any already summarised
+			if (workinfoid <= dbstatus.newest_workmarker_workinfoid)
+				return NULL;
+		}
+
 		if (confirm_sharesummary) {
 			TXT_TO_BIGINT("workinfoid", transfer_data(i_workinfoid), workinfoid);
 
@@ -2237,27 +2251,19 @@ seconf:
 		if (!i_poolinstance)
 			return strdup(reply);
 
-		TXT_TO_BIGINT("workinfoid", transfer_data(i_workinfoid), workinfoid);
-
-		ok = workinfo_age(conn, workinfoid,
-					transfer_data(i_poolinstance),
-					by, code, inet, cd,
-					&ss_first, &ss_last,
-					&ss_count, &s_count, &s_diff);
-
+		ok = workinfo_age(workinfoid, transfer_data(i_poolinstance),
+				  by, code, inet, cd, &ss_first, &ss_last,
+				  &ss_count, &s_count, &s_diff);
 		if (!ok) {
 			LOGERR("%s(%s) %s.failed.DATA", __func__, cmd, id);
 			return strdup("failed.DATA");
 		} else {
-			/* Don't slow down the reload - do them later
-			 * N.B. this means if you abort/terminate the reload,
-			 * next restart will again go back to the oldest
-			 * unaged sharesummary due to a pool terminate */
+			/* Don't slow down the reload - do them later */
 			if (!reloading) {
 				// Aging is a queued item thus the reply is ignored
-				auto_age_older(conn, workinfoid,
-						     transfer_data(i_poolinstance),
-						     by, code, inet, cd);
+				auto_age_older(workinfoid,
+						transfer_data(i_poolinstance),
+						by, code, inet, cd);
 			}
 		}
 		LOGDEBUG("%s.ok.aged %"PRId64, id, workinfoid);

src/ckdb_data.c

@@ -1764,10 +1764,10 @@ K_ITEM *next_workinfo(int64_t workinfoid, K_TREE_CTX *ctx)
 	return item;
 }
 
-bool workinfo_age(PGconn *conn, int64_t workinfoid, char *poolinstance,
-		  char *by, char *code, char *inet, tv_t *cd,
-		  tv_t *ss_first, tv_t *ss_last, int64_t *ss_count,
-		  int64_t *s_count, int64_t *s_diff)
+// Duplicates during a reload are set to not show messages
+bool workinfo_age(int64_t workinfoid, char *poolinstance, char *by, char *code,
+		  char *inet, tv_t *cd, tv_t *ss_first, tv_t *ss_last,
+		  int64_t *ss_count, int64_t *s_count, int64_t *s_diff)
 {
 	K_ITEM *wi_item, ss_look, *ss_item, s_look, *s_item;
 	K_ITEM *wm_item, *tmp_item;
@@ -1778,7 +1778,7 @@ bool workinfo_age(PGconn *conn, int64_t workinfoid, char *poolinstance,
 	SHARESUMMARY looksharesummary, *sharesummary;
 	WORKINFO *workinfo;
 	SHARES lookshares, *shares;
-	bool ok = false, conned = false, skipupdate;
+	bool ok = false, skipupdate;
 	char error[1024];
 
 	LOGDEBUG("%s(): age", __func__);
@@ -1857,14 +1857,9 @@ bool workinfo_age(PGconn *conn, int64_t workinfoid, char *poolinstance,
 		}
 
 		if (!skipupdate) {
-			if (conn == NULL && !confirm_sharesummary) {
-				conn = dbconnect();
-				conned = true;
-			}
-
-			if (!sharesummary_update(conn, NULL, NULL, ss_item, by, code, inet, cd)) {
+			if (!sharesummary_update(NULL, NULL, ss_item, by, code, inet, cd)) {
 				ss_failed++;
-				LOGERR("%s(): Failed to age share summary %"PRId64"/%s/%"PRId64,
+				LOGERR("%s(): Failed to age sharesummary %"PRId64"/%s/%"PRId64,
 					__func__, sharesummary->userid,
 					sharesummary->workername,
 					sharesummary->workinfoid);
@@ -1929,9 +1924,6 @@ bool workinfo_age(PGconn *conn, int64_t workinfoid, char *poolinstance,
 			LOGERR("%s(): %s", __func__, error);
 	}
 
-	if (conned)
-		PQfinish(conn);
-
 	if (ss_already || ss_failed || shares_dumped) {
 		/* If all were already aged, and no shares
 		 * then we don't want a message */
@@ -2046,8 +2038,7 @@ void zero_sharesummary(SHARESUMMARY *row, tv_t *cd, double diff)
 	row->diffacc = row->diffsta = row->diffdup = row->diffhi =
 	row->diffrej = row->shareacc = row->sharesta = row->sharedup =
 	row->sharehi = row->sharerej = 0.0;
-	row->sharecount = row->errorcount = row->countlastupdate = 0;
-	row->reset = false;
+	row->sharecount = row->errorcount = 0;
 	row->firstshare.tv_sec = cd->tv_sec;
 	row->firstshare.tv_usec = cd->tv_usec;
 	row->lastshare.tv_sec = row->firstshare.tv_sec;
@@ -2102,8 +2093,8 @@ K_ITEM *find_last_sharesummary(int64_t userid, char *workername)
 
 /* TODO: markersummary checking?
  * However, there should be no issues since the sharesummaries are removed */
-void auto_age_older(PGconn *conn, int64_t workinfoid, char *poolinstance,
-		    char *by, char *code, char *inet, tv_t *cd)
+void auto_age_older(int64_t workinfoid, char *poolinstance, char *by,
+		    char *code, char *inet, tv_t *cd)
 {
 	static int64_t last_attempted_id = -1;
 	static int64_t prev_found = 0;
@@ -2170,10 +2161,9 @@ void auto_age_older(PGconn *conn, int64_t workinfoid, char *poolinstance,
 		do_id = age_id;
 		to_id = 0;
 		do {
-			ok = workinfo_age(conn, do_id, poolinstance,
-						by, code, inet, cd,
-						&ss_first, &ss_last,
-						&ss_count, &s_count, &s_diff);
+			ok = workinfo_age(do_id, poolinstance, by, code, inet,
+					  cd, &ss_first, &ss_last, &ss_count,
+					  &s_count, &s_diff);
 
 			ss_count_tot += ss_count;
 			s_count_tot += s_count;
@@ -4611,5 +4601,5 @@ void _userinfo_block(BLOCKS *blocks, bool isnew, bool lock)
 	} else
 		row->orphans++;
 	if (lock)
-		K_WLOCK(userinfo_free);
+		K_WUNLOCK(userinfo_free);
 }

src/ckdb_dbio.c

@@ -2760,8 +2760,10 @@ bool workinfo_fill(PGconn *conn)
 			workinfo_height_root = add_to_ktree(workinfo_height_root, item, cmp_workinfo_height);
 		k_add_head(workinfo_store, item);
 
-		if (tv_newer(&(dbstatus.newest_createdate_workinfo), &(row->createdate)))
+		if (tv_newer(&(dbstatus.newest_createdate_workinfo), &(row->createdate))) {
 			copy_tv(&(dbstatus.newest_createdate_workinfo), &(row->createdate));
+			dbstatus.newest_workinfoid = row->workinfoid;
+		}
 
 		tick();
 	}
@@ -2834,16 +2836,6 @@ static bool shares_process(PGconn *conn, SHARES *shares, K_TREE *trf_root)
 				// Reloading a share already summarised
 				return true;
 			}
-
-			if (!sharesummary->reset) {
-				_userinfo_update(NULL, sharesummary, NULL,
-						 true, true);
-
-				zero_sharesummary(sharesummary,
-						  &(shares->createdate),
-						  shares->diff);
-				sharesummary->reset = true;
-			}
 		}
 	}
 
@@ -2852,7 +2844,7 @@ static bool shares_process(PGconn *conn, SHARES *shares, K_TREE *trf_root)
 		userinfo_update(shares, NULL, NULL);
 	}
 
-	sharesummary_update(conn, shares, NULL, NULL, shares->createby,
+	sharesummary_update(shares, NULL, NULL, shares->createby,
 			    shares->createcode, shares->createinet,
 			    &(shares->createdate));
 
@@ -3147,20 +3139,10 @@ static bool shareerrors_process(PGconn *conn, SHAREERRORS *shareerrors,
 				FREENULL(st);
 				return false;
 			}
-
-			if (!sharesummary->reset) {
-				_userinfo_update(NULL, sharesummary, NULL,
-						 true, true);
-
-				zero_sharesummary(sharesummary,
-						  &(shareerrors->createdate),
-						  0.0);
-				sharesummary->reset = true;
-			}
 		}
 	}
 
-	sharesummary_update(conn, NULL, shareerrors, NULL,
+	sharesummary_update(NULL, shareerrors, NULL,
 			    shareerrors->createby,
 			    shareerrors->createcode,
 			    shareerrors->createinet,
@@ -3768,6 +3750,8 @@ flail:
 	return ok;
 }
 
+// no longer used
+#if 0
 static void sharesummary_to_pool(SHARESUMMARY *p_row, SHARESUMMARY *row)
 {
 	p_row->diffacc += row->diffacc;
@@ -3791,6 +3775,7 @@ static void sharesummary_to_pool(SHARESUMMARY *p_row, SHARESUMMARY *row)
 		p_row->lastdiffacc = row->lastdiffacc;
 	}
 }
+#endif
 
 static void set_sharesummary_stats(SHARESUMMARY *row, SHARES *s_row,
 				   SHAREERRORS *e_row, bool new,
@@ -3866,22 +3851,18 @@ char *ooo_status(char *buf, size_t siz)
 	return buf;
 }
 
-bool _sharesummary_update(PGconn *conn, SHARES *s_row, SHAREERRORS *e_row, K_ITEM *ss_item,
-				char *by, char *code, char *inet, tv_t *cd, WHERE_FFL_ARGS)
+// No longer stored in the DB but fields are updated as before
+bool _sharesummary_update(SHARES *s_row, SHAREERRORS *e_row, K_ITEM *ss_item,
+				char *by, char *code, char *inet, tv_t *cd,
+				WHERE_FFL_ARGS)
 {
-	ExecStatusType rescode;
-	PGresult *res = NULL;
 	WORKMARKERS *wm;
 	SHARESUMMARY *row, *p_row;
 	K_ITEM *item, *wm_item, *p_item = NULL;
-	char *ins, *upd;
-	bool ok = false, new = false, p_new = false;
-	char *params[19 + MODIFYDATECOUNT];
-	int n, par = 0;
+	bool new = false, p_new = false;
 	int64_t userid, workinfoid;
 	char *workername;
 	tv_t *createdate;
-	bool must_update = false, conned = false;
 	char *st = NULL, *db = NULL;
 	char ooo_buf[256];
 	double tdf, tdl;
@@ -3897,7 +3878,6 @@ bool _sharesummary_update(PGconn *conn, SHARES *s_row, SHAREERRORS *e_row, K_ITE
 		}
 		item = ss_item;
 		DATA_SHARESUMMARY(row, item);
-		must_update = true;
 		row->complete[0] = SUMMARY_COMPLETE;
 		row->complete[1] = '\0';
 	} else {
@@ -3958,8 +3938,6 @@ bool _sharesummary_update(PGconn *conn, SHARES *s_row, SHAREERRORS *e_row, K_ITE
 			row->workername = strdup(workername);
 			LIST_MEM_ADD(sharesummary_free, row->workername);
 			row->workinfoid = workinfoid;
-			row->inserted = false;
-			row->saveaged = false;
 		}
 
 		// N.B. this directly updates the non-key data
@@ -4018,7 +3996,6 @@ bool _sharesummary_update(PGconn *conn, SHARES *s_row, SHAREERRORS *e_row, K_ITE
 			}
 		}
 
-		// p_items are ram only
 		if (p_item) {
 			DATA_SHARESUMMARY(p_row, p_item);
 		} else {
@@ -4036,161 +4013,9 @@ bool _sharesummary_update(PGconn *conn, SHARES *s_row, SHAREERRORS *e_row, K_ITE
 		set_sharesummary_stats(p_row, s_row, e_row, p_new, &tdf, &tdl);
 	}
 
-	// During startup, don't save 'new' sharesummaries, to reduce DB I/O
-	// ... and also during normal processing
-	if (row->complete[0] == SUMMARY_NEW)
-		goto startupskip;
-
-	if (conn == NULL && !confirm_sharesummary) {
-		conn = dbconnect();
-		conned = true;
-	}
-
-	if (new || !(row->inserted)) {
-		MODIFYDATEPOINTERS(sharesummary_free, row, cd, by, code, inet);
-
-		if (!confirm_sharesummary) {
-			par = 0;
-			params[par++] = bigint_to_buf(row->userid, NULL, 0);
-			params[par++] = str_to_buf(row->workername, NULL, 0);
-			params[par++] = bigint_to_buf(row->workinfoid, NULL, 0);
-			params[par++] = double_to_buf(row->diffacc, NULL, 0);
-			params[par++] = double_to_buf(row->diffsta, NULL, 0);
-			params[par++] = double_to_buf(row->diffdup, NULL, 0);
-			params[par++] = double_to_buf(row->diffhi, NULL, 0);
-			params[par++] = double_to_buf(row->diffrej, NULL, 0);
-			params[par++] = double_to_buf(row->shareacc, NULL, 0);
-			params[par++] = double_to_buf(row->sharesta, NULL, 0);
-			params[par++] = double_to_buf(row->sharedup, NULL, 0);
-			params[par++] = double_to_buf(row->sharehi, NULL, 0);
-			params[par++] = double_to_buf(row->sharerej, NULL, 0);
-			params[par++] = bigint_to_buf(row->sharecount, NULL, 0);
-			params[par++] = bigint_to_buf(row->errorcount, NULL, 0);
-			params[par++] = tv_to_buf(&(row->firstshare), NULL, 0);
-			params[par++] = tv_to_buf(&(row->lastshare), NULL, 0);
-			params[par++] = double_to_buf(row->lastdiffacc, NULL, 0);
-			params[par++] = str_to_buf(row->complete, NULL, 0);
-			MODIFYDATEPARAMS(params, par, row);
-			PARCHK(par, params);
-
-			ins = "insert into sharesummary "
-				"(userid,workername,workinfoid,diffacc,diffsta,diffdup,diffhi,"
-				"diffrej,shareacc,sharesta,sharedup,sharehi,sharerej,"
-				"sharecount,errorcount,firstshare,lastshare,"
-				"lastdiffacc,complete"
-				MODIFYDATECONTROL ") values (" PQPARAM27 ")";
-
-			res = PQexecParams(conn, ins, par, NULL, (const char **)params, NULL, NULL, 0, CKPQ_WRITE);
-			rescode = PQresultStatus(res);
-			if (!PGOK(rescode)) {
-				PGLOGERR("Insert", rescode, conn);
-				goto unparam;
-			}
-		}
-
-		row->countlastupdate = row->sharecount + row->errorcount;
-		row->inserted = true;
-		if (row->complete[0] == SUMMARY_COMPLETE)
-			row->saveaged = true;
-	} else {
-		bool stats_update = false;
-
-		MODIFYUPDATEPOINTERS(sharesummary_free, row, cd, by, code, inet);
-
-		if ((row->countlastupdate + SHARESUMMARY_UPDATE_EVERY) <
-		    (row->sharecount + row->errorcount))
-			stats_update = true;
-
-		if (must_update && row->countlastupdate < (row->sharecount + row->errorcount))
-			stats_update = true;
+	MODIFYDATEPOINTERS(sharesummary_free, row, cd, by, code, inet);
 
-		if (stats_update) {
-			if (!confirm_sharesummary) {
-				par = 0;
-				params[par++] = bigint_to_buf(row->userid, NULL, 0);
-				params[par++] = str_to_buf(row->workername, NULL, 0);
-				params[par++] = bigint_to_buf(row->workinfoid, NULL, 0);
-				params[par++] = double_to_buf(row->diffacc, NULL, 0);
-				params[par++] = double_to_buf(row->diffsta, NULL, 0);
-				params[par++] = double_to_buf(row->diffdup, NULL, 0);
-				params[par++] = double_to_buf(row->diffhi, NULL, 0);
-				params[par++] = double_to_buf(row->diffrej, NULL, 0);
-				params[par++] = double_to_buf(row->shareacc, NULL, 0);
-				params[par++] = double_to_buf(row->sharesta, NULL, 0);
-				params[par++] = double_to_buf(row->sharedup, NULL, 0);
-				params[par++] = double_to_buf(row->sharehi, NULL, 0);
-				params[par++] = double_to_buf(row->sharerej, NULL, 0);
-				params[par++] = tv_to_buf(&(row->firstshare), NULL, 0);
-				params[par++] = tv_to_buf(&(row->lastshare), NULL, 0);
-				params[par++] = bigint_to_buf(row->sharecount, NULL, 0);
-				params[par++] = bigint_to_buf(row->errorcount, NULL, 0);
-				params[par++] = double_to_buf(row->lastdiffacc, NULL, 0);
-				params[par++] = str_to_buf(row->complete, NULL, 0);
-				MODIFYUPDATEPARAMS(params, par, row);
-				PARCHKVAL(par, 23, params);
-
-				upd = "update sharesummary "
-					"set diffacc=$4,diffsta=$5,diffdup=$6,diffhi=$7,diffrej=$8,"
-					"shareacc=$9,sharesta=$10,sharedup=$11,sharehi=$12,"
-					"sharerej=$13,firstshare=$14,lastshare=$15,"
-					"sharecount=$16,errorcount=$17,lastdiffacc=$18,complete=$19"
-					","MDDB"=$20,"MBYDB"=$21,"MCODEDB"=$22,"MINETDB"=$23 "
-					"where userid=$1 and workername=$2 and workinfoid=$3";
-
-				res = PQexecParams(conn, upd, par, NULL, (const char **)params, NULL, NULL, 0, CKPQ_WRITE);
-				rescode = PQresultStatus(res);
-				if (!PGOK(rescode)) {
-					PGLOGERR("Update", rescode, conn);
-					goto unparam;
-				}
-			}
-			row->countlastupdate = row->sharecount + row->errorcount;
-			if (row->complete[0] == SUMMARY_COMPLETE)
-				row->saveaged = true;
-		} else {
-			if (!must_update) {
-				ok = true;
-				goto late;
-			} else {
-				if (!confirm_sharesummary) {
-					par = 0;
-					params[par++] = bigint_to_buf(row->userid, NULL, 0);
-					params[par++] = str_to_buf(row->workername, NULL, 0);
-					params[par++] = bigint_to_buf(row->workinfoid, NULL, 0);
-					params[par++] = str_to_buf(row->complete, NULL, 0);
-					MODIFYUPDATEPARAMS(params, par, row);
-					PARCHKVAL(par, 8, params);
-
-					upd = "update sharesummary "
-						"set complete=$4,"MDDB"=$5,"MBYDB"=$6,"MCODEDB"=$7,"MINETDB"=$8 "
-						"where userid=$1 and workername=$2 and workinfoid=$3";
-
-					res = PQexecParams(conn, upd, par, NULL, (const char **)params, NULL, NULL, 0, CKPQ_WRITE);
-					rescode = PQresultStatus(res);
-					if (!PGOK(rescode)) {
-						PGLOGERR("MustUpdate", rescode, conn);
-						goto unparam;
-					}
-				}
-				row->countlastupdate = row->sharecount + row->errorcount;
-				if (row->complete[0] == SUMMARY_COMPLETE)
-					row->saveaged = true;
-			}
-		}
-	}
-startupskip:
-	ok = true;
-unparam:
-	if (par) {
-		PQclear(res);
-		for (n = 0; n < par; n++)
-			free(params[n]);
-	}
-late:
-	if (conned)
-		PQfinish(conn);
-
-	// We keep the new item no matter what 'ok' is, since it will be inserted later
+	// Store either new item
 	if (new || p_new) {
 		K_WLOCK(sharesummary_free);
 		if (new) {
@@ -4209,267 +4034,7 @@ late:
 		K_WUNLOCK(sharesummary_free);
 	}
 
-	return ok;
-}
-
-bool sharesummary_fill(PGconn *conn)
-{
-	ExecStatusType rescode;
-	PGresult *res;
-	K_TREE_CTX ctx[1];
-	K_ITEM *item, *m_item, *p_item;
-	int n, i, par = 0, p_n;
-	SHARESUMMARY *row, *p_row;
-	MARKS *marks;
-	char *params[2];
-	char *field;
-	char *sel;
-	int fields = 19;
-	bool ok;
-
-	LOGDEBUG("%s(): select", __func__);
-
-	/* Load needs to go back to the last marks workinfoid(+1)
-	 * If it is later than that, we can't create markersummaries
-	 *  since some of the required data is missing -
-	 *  thus we also can't make the shift markersummaries */
-	m_item = last_in_ktree(marks_root, ctx);
-	if (!m_item) {
-		if (dbload_workinfoid_start != -1) {
-			sharesummary_marks_limit = true;
-			LOGWARNING("WARNING: dbload -w start used "
-				   "but there are no marks ...");
-		}
-	} else {
-		DATA_MARKS(marks, m_item);
-		if (dbload_workinfoid_start > marks->workinfoid) {
-			sharesummary_marks_limit = true;
-			LOGWARNING("WARNING: dbload -w start %"PRId64
-				   " is after the last mark %"PRId64" ...",
-				   dbload_workinfoid_start,
-				   marks->workinfoid);
-		}
-	}
-	if (sharesummary_marks_limit) {
-		LOGWARNING("WARNING: ... markersummaries cannot be created "
-			   "and pplns calculations may be wrong");
-	}
-
-	sel = "select "
-		"userid,workername,workinfoid,diffacc,diffsta,diffdup,diffhi,"
-		"diffrej,shareacc,sharesta,sharedup,sharehi,sharerej,"
-		"sharecount,errorcount,firstshare,lastshare,"
-		"lastdiffacc,complete"
-		MODIFYDATECONTROL
-		" from sharesummary where workinfoid>=$1 and workinfoid<=$2";
-	par = 0;
-	params[par++] = bigint_to_buf(dbload_workinfoid_start, NULL, 0);
-	params[par++] = bigint_to_buf(dbload_workinfoid_finish, NULL, 0);
-	PARCHK(par, params);
-	res = PQexecParams(conn, sel, par, NULL, (const char **)params, NULL, NULL, 0, CKPQ_READ);
-	rescode = PQresultStatus(res);
-	if (!PGOK(rescode)) {
-		PGLOGERR("Select", rescode, conn);
-		PQclear(res);
-		return false;
-	}
-
-	n = PQnfields(res);
-	if (n != (fields + MODIFYDATECOUNT)) {
-		LOGERR("%s(): Invalid field count - should be %d, but is %d",
-			__func__, fields + MODIFYDATECOUNT, n);
-		PQclear(res);
-		return false;
-	}
-
-	n = PQntuples(res);
-	LOGDEBUG("%s(): tree build count %d", __func__, n);
-	ok = true;
-	//K_WLOCK(sharesummary_free);
-	for (i = 0; i < n; i++) {
-		item = k_unlink_head(sharesummary_free);
-		DATA_SHARESUMMARY(row, item);
-		bzero(row, sizeof(*row));
-
-		if (everyone_die) {
-			ok = false;
-			break;
-		}
-
-		row->inserted = true;
-
-		PQ_GET_FLD(res, i, "userid", field, ok);
-		if (!ok)
-			break;
-		TXT_TO_BIGINT("userid", field, row->userid);
-
-		PQ_GET_FLD(res, i, "workername", field, ok);
-		if (!ok)
-			break;
-		TXT_TO_PTR("workername", field, row->workername);
-		LIST_MEM_ADD(sharesummary_free, row->workername);
-
-		PQ_GET_FLD(res, i, "workinfoid", field, ok);
-		if (!ok)
-			break;
-		TXT_TO_BIGINT("workinfoid", field, row->workinfoid);
-
-		PQ_GET_FLD(res, i, "diffacc", field, ok);
-		if (!ok)
-			break;
-		TXT_TO_DOUBLE("diffacc", field, row->diffacc);
-
-		PQ_GET_FLD(res, i, "diffsta", field, ok);
-		if (!ok)
-			break;
-		TXT_TO_DOUBLE("diffsta", field, row->diffsta);
-
-		PQ_GET_FLD(res, i, "diffdup", field, ok);
-		if (!ok)
-			break;
-		TXT_TO_DOUBLE("diffdup", field, row->diffdup);
-
-		PQ_GET_FLD(res, i, "diffhi", field, ok);
-		if (!ok)
-			break;
-		TXT_TO_DOUBLE("diffhi", field, row->diffhi);
-
-		PQ_GET_FLD(res, i, "diffrej", field, ok);
-		if (!ok)
-			break;
-		TXT_TO_DOUBLE("diffrej", field, row->diffrej);
-
-		PQ_GET_FLD(res, i, "shareacc", field, ok);
-		if (!ok)
-			break;
-		TXT_TO_DOUBLE("shareacc", field, row->shareacc);
-
-		PQ_GET_FLD(res, i, "sharesta", field, ok);
-		if (!ok)
-			break;
-		TXT_TO_DOUBLE("sharesta", field, row->sharesta);
-
-		PQ_GET_FLD(res, i, "sharedup", field, ok);
-		if (!ok)
-			break;
-		TXT_TO_DOUBLE("sharedup", field, row->sharedup);
-
-		PQ_GET_FLD(res, i, "sharehi", field, ok);
-		if (!ok)
-			break;
-		TXT_TO_DOUBLE("sharehi", field, row->sharehi);
-
-		PQ_GET_FLD(res, i, "sharerej", field, ok);
-		if (!ok)
-			break;
-		TXT_TO_DOUBLE("sharerej", field, row->sharerej);
-
-		PQ_GET_FLD(res, i, "sharecount", field, ok);
-		if (!ok)
-			break;
-		TXT_TO_BIGINT("sharecount", field, row->sharecount);
-
-		PQ_GET_FLD(res, i, "errorcount", field, ok);
-		if (!ok)
-			break;
-		TXT_TO_BIGINT("errorcount", field, row->errorcount);
-
-		row->countlastupdate = row->sharecount + row->errorcount;
-
-		PQ_GET_FLD(res, i, "firstshare", field, ok);
-		if (!ok)
-			break;
-		TXT_TO_TV("firstshare", field, row->firstshare);
-
-		PQ_GET_FLD(res, i, "lastshare", field, ok);
-		if (!ok)
-			break;
-		TXT_TO_TV("lastshare", field, row->lastshare);
-
-		PQ_GET_FLD(res, i, "lastdiffacc", field, ok);
-		if (!ok)
-			break;
-		TXT_TO_DOUBLE("lastdiffacc", field, row->lastdiffacc);
-
-		PQ_GET_FLD(res, i, "complete", field, ok);
-		if (!ok)
-			break;
-		TXT_TO_STR("complete", field, row->complete);
-
-		MODIFYDATEFLDPOINTERS(sharesummary_free, res, i, row, ok);
-		if (!ok)
-			break;
-
-		sharesummary_root = add_to_ktree(sharesummary_root, item, cmp_sharesummary);
-		sharesummary_workinfoid_root = add_to_ktree(sharesummary_workinfoid_root, item, cmp_sharesummary_workinfoid);
-		k_add_head(sharesummary_store, item);
-
-		// A share summary is shares in a single workinfo, at all 3 levels n,a,y
-		if (tolower(row->complete[0]) == SUMMARY_NEW) {
-			if (dbstatus.oldest_sharesummary_firstshare_n.tv_sec == 0 ||
-			    !tv_newer(&(dbstatus.oldest_sharesummary_firstshare_n), &(row->firstshare))) {
-				copy_tv(&(dbstatus.oldest_sharesummary_firstshare_n), &(row->firstshare));
-				dbstatus.oldest_workinfoid_n = row->workinfoid;
-			}
-		} else {
-			if (tv_newer(&(dbstatus.newest_sharesummary_firstshare_ay), &(row->firstshare)))
-				copy_tv(&(dbstatus.newest_sharesummary_firstshare_ay), &(row->firstshare));
-			if (tolower(row->complete[0]) == SUMMARY_COMPLETE) {
-				if (dbstatus.oldest_sharesummary_firstshare_a.tv_sec == 0 ||
-				    !tv_newer(&(dbstatus.oldest_sharesummary_firstshare_a), &(row->firstshare))) {
-					copy_tv(&(dbstatus.oldest_sharesummary_firstshare_a), &(row->firstshare));
-					dbstatus.oldest_workinfoid_a = row->workinfoid;
-				}
-				if (tv_newer(&(dbstatus.newest_sharesummary_firstshare_a), &(row->firstshare))) {
-					copy_tv(&(dbstatus.newest_sharesummary_firstshare_a), &(row->firstshare));
-					dbstatus.newest_workinfoid_a = row->workinfoid;
-				}
-			} else /* SUMMARY_CONFIRM */ {
-				if (tv_newer(&(dbstatus.newest_sharesummary_firstshare_y), &(row->firstshare))) {
-					copy_tv(&(dbstatus.newest_sharesummary_firstshare_y), &(row->firstshare));
-					dbstatus.newest_workinfoid_y = row->workinfoid;
-				}
-			}
-		}
-
-		p_item = find_sharesummary_p(row->workinfoid);
-		if (!p_item) {
-			p_item = k_unlink_head(sharesummary_free);
-			DATA_SHARESUMMARY(p_row, p_item);
-			bzero(p_row, sizeof(*p_row));
-			POOL_SS(p_row);
-			LIST_MEM_ADD(sharesummary_free, p_row->workername);
-			p_row->workinfoid = row->workinfoid;
-			sharesummary_pool_root = add_to_ktree(sharesummary_pool_root,
-							      p_item,
-							      cmp_sharesummary);
-			k_add_head(sharesummary_pool_store, p_item);
-		} else {
-			DATA_SHARESUMMARY(p_row, p_item);
-		}
-
-		sharesummary_to_pool(p_row, row);
-
-		_userinfo_update(NULL, row, NULL, false, false);
-
-		tick();
-	}
-	if (!ok) {
-		FREENULL(row->workername);
-		k_add_head(sharesummary_free, item);
-	}
-
-	p_n = sharesummary_pool_store->count;
-	//K_WUNLOCK(sharesummary_free);
-	PQclear(res);
-
-	if (ok) {
-		LOGDEBUG("%s(): built", __func__);
-		LOGWARNING("%s(): loaded %d sharesummary records", __func__, n);
-		LOGWARNING("%s(): created %d sharesummary pool records", __func__, p_n);
-	}
-
-	return ok;
+	return true;
 }
 
 bool blocks_stats(PGconn *conn, int32_t height, char *blockhash,
@@ -5855,7 +5420,15 @@ bool poolstats_add(PGconn *conn, bool store, char *poolinstance,
 		res = PQexecParams(conn, ins, par, NULL, (const char **)params, NULL, NULL, 0, CKPQ_WRITE);
 		rescode = PQresultStatus(res);
 		if (!PGOK(rescode)) {
-			PGLOGERR("Insert", rescode, conn);
+			bool show_msg = true;
+			char *code;
+			if (igndup) {
+				code = PQresultErrorField(res, PG_DIAG_SQLSTATE);
+				if (code && strcmp(code, SQL_UNIQUE_VIOLATION) == 0)
+					show_msg = false;
+			}
+			if (show_msg)
+				PGLOGERR("Insert", rescode, conn);
 			goto unparam;
 		}
 
@@ -6750,7 +6323,8 @@ bool workmarkers_fill(PGconn *conn)
 {
 	ExecStatusType rescode;
 	PGresult *res;
-	K_ITEM *item;
+	K_ITEM *item, *wi_item;
+	WORKINFO *workinfo;
 	int n, i;
 	WORKMARKERS *row;
 	char *field;
@@ -6837,6 +6411,23 @@ bool workmarkers_fill(PGconn *conn)
 							   item, cmp_workmarkers_workinfoid);
 		k_add_head(workmarkers_store, item);
 
+		if (dbstatus.newest_workmarker_workinfoid < row->workinfoidend) {
+			dbstatus.newest_workmarker_workinfoid = row->workinfoidend;
+			wi_item = find_workinfo(row->workinfoidend, NULL);
+			if (!wi_item) {
+				LOGEMERG("%s(): FAILURE workmarkerid %"PRId64
+					 " wid end %"PRId64" doesn't exist! "
+					 "You should abort ckdb and fix it, "
+					 " since the reload may skip some data",
+					 __func__, row->markerid,
+					 row->workinfoidend);
+			} else {
+				DATA_WORKINFO(workinfo, wi_item);
+				copy_tv(&(dbstatus.newest_createdate_workmarker_workinfo),
+					&(workinfo->createdate));
+			}
+		}
+
 		tick();
 	}
 	if (!ok)