+// This file contains the library routines for managing the
+// file_summary_by_instance table.
+package file_summary_by_instance
+
+import (
+ "database/sql"
+ "fmt"
+ "log"
+ "regexp"
+ "sort"
+ "time"
+
+ "github.com/sjmudd/pstop/key_value_cache"
+ "github.com/sjmudd/pstop/lib"
+)
+
+/*
+CREATE TABLE `file_summary_by_instance` (
+ `FILE_NAME` varchar(512) NOT NULL,
+ `EVENT_NAME` varchar(128) NOT NULL, // not collected
+ `OBJECT_INSTANCE_BEGIN` bigint(20) unsigned NOT NULL, // not collected
+ `COUNT_STAR` bigint(20) unsigned NOT NULL,
+ `SUM_TIMER_WAIT` bigint(20) unsigned NOT NULL,
+ `MIN_TIMER_WAIT` bigint(20) unsigned NOT NULL,
+ `AVG_TIMER_WAIT` bigint(20) unsigned NOT NULL,
+ `MAX_TIMER_WAIT` bigint(20) unsigned NOT NULL,
+ `COUNT_READ` bigint(20) unsigned NOT NULL,
+ `SUM_TIMER_READ` bigint(20) unsigned NOT NULL,
+ `MIN_TIMER_READ` bigint(20) unsigned NOT NULL,
+ `AVG_TIMER_READ` bigint(20) unsigned NOT NULL,
+ `MAX_TIMER_READ` bigint(20) unsigned NOT NULL,
+ `SUM_NUMBER_OF_BYTES_READ` bigint(20) NOT NULL,
+ `COUNT_WRITE` bigint(20) unsigned NOT NULL,
+ `SUM_TIMER_WRITE` bigint(20) unsigned NOT NULL,
+ `MIN_TIMER_WRITE` bigint(20) unsigned NOT NULL,
+ `AVG_TIMER_WRITE` bigint(20) unsigned NOT NULL,
+ `MAX_TIMER_WRITE` bigint(20) unsigned NOT NULL,
+ `SUM_NUMBER_OF_BYTES_WRITE` bigint(20) NOT NULL,
+ `COUNT_MISC` bigint(20) unsigned NOT NULL,
+ `SUM_TIMER_MISC` bigint(20) unsigned NOT NULL,
+ `MIN_TIMER_MISC` bigint(20) unsigned NOT NULL,
+ `AVG_TIMER_MISC` bigint(20) unsigned NOT NULL,
+ `MAX_TIMER_MISC` bigint(20) unsigned NOT NULL
+) ENGINE=PERFORMANCE_SCHEMA DEFAULT CHARSET=utf8
+1 row in set (0.00 sec)
+*/
+
+// foo/../bar --> foo/bar perl: $new =~ s{[^/]+/\.\./}{/};
+// /./ --> / perl: $new =~ s{/\./}{};
+// // --> / perl: $new =~ s{//}{/};
+const (
+ re_encoded = `@(\d{4})` // FIXME - add me to catch @0024 --> $ for example
+)
+
+var (
+ re_one_or_the_other *regexp.Regexp = regexp.MustCompile(`/(\.)?/`)
+ re_slash_dot_dot_slash *regexp.Regexp = regexp.MustCompile(`[^/]+/\.\./`)
+ re_table_file *regexp.Regexp = regexp.MustCompile(`/([^/]+)/([^/]+)\.(frm|ibd|MYD|MYI|CSM|CSV|par)$`)
+ re_temp_table *regexp.Regexp = regexp.MustCompile(`#sql-[0-9_]+`)
+ re_part_table *regexp.Regexp = regexp.MustCompile(`(.+)#P#p(\d+|MAX)`)
+ re_ibdata *regexp.Regexp = regexp.MustCompile(`/ibdata\d+$`)
+ re_redo_log *regexp.Regexp = regexp.MustCompile(`/ib_logfile\d+$`)
+ re_binlog *regexp.Regexp = regexp.MustCompile(`/binlog\.(\d{6}|index)$`)
+ re_db_opt *regexp.Regexp = regexp.MustCompile(`/db\.opt$`)
+ re_slowlog *regexp.Regexp = regexp.MustCompile(`/slowlog$`)
+ re_auto_cnf *regexp.Regexp = regexp.MustCompile(`/auto\.cnf$`)
+ re_pid_file *regexp.Regexp = regexp.MustCompile(`/[^/]+\.pid$`)
+ re_error_msg *regexp.Regexp = regexp.MustCompile(`/share/[^/]+/errmsg\.sys$`)
+ re_charset *regexp.Regexp = regexp.MustCompile(`/share/charsets/Index\.xml$`)
+ re_dollar *regexp.Regexp = regexp.MustCompile(`@0024`) // FIXME - add me to catch @0024 --> $ (specific case)
+
+ cache key_value_cache.KeyValueCache
+)
+
+type file_summary_by_instance_row struct {
+ FILE_NAME string
+
+ COUNT_STAR uint64
+ COUNT_READ uint64
+ COUNT_WRITE uint64
+ COUNT_MISC uint64
+
+ SUM_TIMER_WAIT uint64
+ SUM_TIMER_READ uint64
+ SUM_TIMER_WRITE uint64
+ SUM_TIMER_MISC uint64
+
+ SUM_NUMBER_OF_BYTES_READ uint64
+ SUM_NUMBER_OF_BYTES_WRITE uint64
+}
+
+// represents a table or set of rows
+type file_summary_by_instance_rows []file_summary_by_instance_row
+
+// Return the name using the FILE_NAME attribute.
+func (r *file_summary_by_instance_row) name() string {
+ return r.FILE_NAME
+}
+
+// Return a formatted pretty name for the row.
+func (r *file_summary_by_instance_row) pretty_name() string {
+ s := r.name()
+ if len(s) > 30 {
+ s = s[:29]
+ }
+ return fmt.Sprintf("%-30s", s)
+}
+
+func (r *file_summary_by_instance_row) headings() string {
+ return fmt.Sprintf("%-30s %10s %6s|%6s %6s %6s|%8s %8s|%8s %6s %6s %6s",
+ "Table Name",
+ "Latency",
+ "%",
+ "Read",
+ "Write",
+ "Misc",
+ "Rd bytes",
+ "Wr bytes",
+ "Ops",
+ "R Ops",
+ "W Ops",
+ "M Ops")
+}
+
+// generate a printable result
+func (row *file_summary_by_instance_row) row_content(totals file_summary_by_instance_row) string {
+ var name string
+
+ // We assume that if COUNT_STAR = 0 then there's no data at all...
+ // when we have no data we really don't want to show the name either.
+ if row.COUNT_STAR == 0 {
+ name = ""
+ } else {
+ name = row.pretty_name()
+ }
+
+ return fmt.Sprintf("%-30s %10s %6s|%6s %6s %6s|%8s %8s|%8s %6s %6s %6s",
+ name,
+ lib.FormatTime(row.SUM_TIMER_WAIT),
+ lib.FormatPct(lib.MyDivide(row.SUM_TIMER_WAIT, totals.SUM_TIMER_WAIT)),
+ lib.FormatPct(lib.MyDivide(row.SUM_TIMER_READ, row.SUM_TIMER_WAIT)),
+ lib.FormatPct(lib.MyDivide(row.SUM_TIMER_WRITE, row.SUM_TIMER_WAIT)),
+ lib.FormatPct(lib.MyDivide(row.SUM_TIMER_MISC, row.SUM_TIMER_WAIT)),
+ lib.FormatAmount(row.SUM_NUMBER_OF_BYTES_READ),
+ lib.FormatAmount(row.SUM_NUMBER_OF_BYTES_WRITE),
+ lib.FormatAmount(row.COUNT_STAR),
+ lib.FormatPct(lib.MyDivide(row.COUNT_READ, row.COUNT_STAR)),
+ lib.FormatPct(lib.MyDivide(row.COUNT_WRITE, row.COUNT_STAR)),
+ lib.FormatPct(lib.MyDivide(row.COUNT_MISC, row.COUNT_STAR)))
+}
+
+func (this *file_summary_by_instance_row) add(other file_summary_by_instance_row) {
+ this.COUNT_STAR += other.COUNT_STAR
+ this.COUNT_READ += other.COUNT_READ
+ this.COUNT_WRITE += other.COUNT_WRITE
+ this.COUNT_MISC += other.COUNT_MISC
+
+ this.SUM_TIMER_WAIT += other.SUM_TIMER_WAIT
+ this.SUM_TIMER_READ += other.SUM_TIMER_READ
+ this.SUM_TIMER_WRITE += other.SUM_TIMER_WRITE
+ this.SUM_TIMER_MISC += other.SUM_TIMER_MISC
+
+ this.SUM_NUMBER_OF_BYTES_READ += other.SUM_NUMBER_OF_BYTES_READ
+ this.SUM_NUMBER_OF_BYTES_WRITE += other.SUM_NUMBER_OF_BYTES_WRITE
+}
+
+func (this *file_summary_by_instance_row) subtract(other file_summary_by_instance_row) {
+ this.COUNT_STAR -= other.COUNT_STAR
+ this.COUNT_READ -= other.COUNT_READ
+ this.COUNT_WRITE -= other.COUNT_WRITE
+ this.COUNT_MISC -= other.COUNT_MISC
+
+ this.SUM_TIMER_WAIT -= other.SUM_TIMER_WAIT
+ this.SUM_TIMER_READ -= other.SUM_TIMER_READ
+ this.SUM_TIMER_WRITE -= other.SUM_TIMER_WRITE
+ this.SUM_TIMER_MISC -= other.SUM_TIMER_MISC
+
+ this.SUM_NUMBER_OF_BYTES_READ -= other.SUM_NUMBER_OF_BYTES_READ
+ this.SUM_NUMBER_OF_BYTES_WRITE -= other.SUM_NUMBER_OF_BYTES_WRITE
+}
+
+// return the totals of a slice of rows
+func (t file_summary_by_instance_rows) totals() file_summary_by_instance_row {
+ var totals file_summary_by_instance_row
+ totals.FILE_NAME = "TOTALS"
+
+ for i := range t {
+ totals.add(t[i])
+ }
+
+ return totals
+}
+
+// clean up the given path reducing redundant stuff and return the clean path
+func cleanup_path(path string) string {
+
+ for {
+ orig_path := path
+ path = re_one_or_the_other.ReplaceAllString(path, "/")
+ path = re_slash_dot_dot_slash.ReplaceAllString(path, "/")
+ if orig_path == path { // no change so give up
+ break
+ }
+ }
+
+ return path
+}
+
+// From the original FILE_NAME we want to generate a simpler name to use.
+// This simpler name may also merge several different filenames into one.
+func (t file_summary_by_instance_row) simple_name(global_variables map[string]string) string {
+
+ path := t.FILE_NAME
+
+ if cached_result, err := cache.Get(path); err == nil {
+ return cached_result
+ }
+
+ // @0024 --> $ (should do this more generically)
+ path = re_dollar.ReplaceAllLiteralString(path, "$")
+
+ // this should probably be ordered from most expected regexp to least
+ if m1 := re_table_file.FindStringSubmatch(path); m1 != nil {
+ // we may match temporary tables so check for them
+ if m2 := re_temp_table.FindStringSubmatch(m1[2]); m2 != nil {
+ return cache.Put(path, "<temp_table>")
+ }
+
+ // we may match partitioned tables so check for them
+ if m3 := re_part_table.FindStringSubmatch(m1[2]); m3 != nil {
+ return cache.Put(path, m1[1]+"."+m3[1]) // <schema>.<table> (less partition info)
+ }
+
+ return cache.Put(path, m1[1]+"."+m1[2]) // <schema>.<table>
+ }
+ if re_ibdata.MatchString(path) == true {
+ return cache.Put(path, "<ibdata>")
+ }
+ if re_redo_log.MatchString(path) == true {
+ return cache.Put(path, "<redo_log>")
+ }
+ if re_binlog.MatchString(path) == true {
+ return cache.Put(path, "<binlog>")
+ }
+ if re_db_opt.MatchString(path) == true {
+ return cache.Put(path, "<db_opt>")
+ }
+ if re_slowlog.MatchString(path) == true {
+ return cache.Put(path, "<slow_log>")
+ }
+ if re_auto_cnf.MatchString(path) == true {
+ return cache.Put(path, "<auto_cnf>")
+ }
+ // relay logs are a bit complicated. If a full path then easy to
+ // identify,but if a relative path we may need to add $datadir,
+ // but also if as I do we have a ../blah/somewhere/path then we
+ // need to make it match too.
+ if len(global_variables["relay_log"]) > 0 {
+ relay_log := global_variables["relay_log"]
+ if relay_log[0] != '/' { // relative path
+ relay_log = cleanup_path(global_variables["datadir"] + relay_log) // datadir always ends in /
+ }
+ re_relay_log := relay_log + `\.(\d{6}|index)$`
+ if regexp.MustCompile(re_relay_log).MatchString(path) == true {
+ return cache.Put(path, "<relay_log>")
+ }
+ }
+ if re_pid_file.MatchString(path) == true {
+ return cache.Put(path, "<pid_file>")
+ }
+ if re_error_msg.MatchString(path) == true {
+ return cache.Put(path, "<errmsg>")
+ }
+ if re_charset.MatchString(path) == true {
+ return cache.Put(path, "<charset>")
+ }
+ // clean up datadir to <datadir>
+ if len(global_variables["datadir"]) > 0 {
+ re_datadir := regexp.MustCompile("^" + global_variables["datadir"])
+ path = re_datadir.ReplaceAllLiteralString(path, "<datadir>/")
+ }
+
+ return cache.Put(path, path)
+}
+
+// Convert the imported "table" to a merged one with merged data.
+// Combine all entries with the same "FILE_NAME" by adding their values.
+func merge_by_table_name(orig file_summary_by_instance_rows, global_variables map[string]string) file_summary_by_instance_rows {
+ start := time.Now()
+ t := make(file_summary_by_instance_rows, 0, len(orig))
+
+ m := make(map[string]file_summary_by_instance_row)
+
+ // iterate over source table
+ for i := range orig {
+ var file_name string
+ var new_row file_summary_by_instance_row
+ orig_row := orig[i]
+
+ if orig_row.COUNT_STAR > 0 {
+ file_name = orig_row.simple_name(global_variables)
+
+ // check if we have an entry in the map
+ if _, found := m[file_name]; found {
+ new_row = m[file_name]
+ } else {
+ new_row.FILE_NAME = file_name
+ }
+ new_row.add(orig_row)
+ m[file_name] = new_row // update the map with the new value
+ }
+ }
+
+ // add the map contents back into the table
+ for _, row := range m {
+ t = append(t, row)
+ }
+
+ lib.Logger.Println("merge_by_table_name() took:", time.Duration(time.Since(start)).String())
+ return t
+}
+
+// Select the raw data from the database into file_summary_by_instance_rows
+// - filter out empty values
+// - merge rows with the same name into a single row
+// - change FILE_NAME into a more descriptive value.
+func select_fsbi_rows(dbh *sql.DB) file_summary_by_instance_rows {
+ var t file_summary_by_instance_rows
+ start := time.Now()
+
+ sql := "SELECT FILE_NAME, COUNT_STAR, SUM_TIMER_WAIT, COUNT_READ, SUM_TIMER_READ, SUM_NUMBER_OF_BYTES_READ, COUNT_WRITE, SUM_TIMER_WRITE, SUM_NUMBER_OF_BYTES_WRITE, COUNT_MISC, SUM_TIMER_MISC FROM file_summary_by_instance"
+
+ rows, err := dbh.Query(sql)
+ if err != nil {
+ log.Fatal(err)
+ }
+ defer rows.Close()
+
+ for rows.Next() {
+ var r file_summary_by_instance_row
+
+ if err := rows.Scan(&r.FILE_NAME, &r.COUNT_STAR, &r.SUM_TIMER_WAIT, &r.COUNT_READ, &r.SUM_TIMER_READ, &r.SUM_NUMBER_OF_BYTES_READ, &r.COUNT_WRITE, &r.SUM_TIMER_WRITE, &r.SUM_NUMBER_OF_BYTES_WRITE, &r.COUNT_MISC, &r.SUM_TIMER_MISC); err != nil {
+ log.Fatal(err)
+ }
+ t = append(t, r)
+ }
+ if err := rows.Err(); err != nil {
+ log.Fatal(err)
+ }
+ lib.Logger.Println("select_fsbi_rows() took:", time.Duration(time.Since(start)).String())
+
+ return t
+}
+
+// remove the initial values from those rows where there's a match
+// - if we find a row we can't match ignore it
+func (this *file_summary_by_instance_rows) subtract(initial file_summary_by_instance_rows) {
+ i_by_name := make(map[string]int)
+
+ // iterate over rows by name
+ for i := range initial {
+ i_by_name[initial[i].name()] = i
+ }
+
+ for i := range *this {
+ if _, ok := i_by_name[(*this)[i].name()]; ok {
+ initial_i := i_by_name[(*this)[i].name()]
+ (*this)[i].subtract(initial[initial_i])
+ }
+ }
+}
+
+func (t file_summary_by_instance_rows) Len() int { return len(t) }
+func (t file_summary_by_instance_rows) Swap(i, j int) { t[i], t[j] = t[j], t[i] }
+func (t file_summary_by_instance_rows) Less(i, j int) bool {
+ return (t[i].SUM_TIMER_WAIT > t[j].SUM_TIMER_WAIT) ||
+ ((t[i].SUM_TIMER_WAIT == t[j].SUM_TIMER_WAIT) && (t[i].FILE_NAME < t[j].FILE_NAME))
+}
+
+func (t *file_summary_by_instance_rows) sort() {
+ sort.Sort(t)
+}
+
+// if the data in t2 is "newer", "has more values" than t then it needs refreshing.
+// check this by comparing totals.
+func (t file_summary_by_instance_rows) needs_refresh(t2 file_summary_by_instance_rows) bool {
+ my_totals := t.totals()
+ t2_totals := t2.totals()
+
+ return my_totals.SUM_TIMER_WAIT > t2_totals.SUM_TIMER_WAIT
+}