richardcrist
commented
2 years ago More research. Not a go expert but took a look into code. Below is a code section ending in the debug display of the value.
// interface method to call ScrapeGenericValues using Metric struct values
func ScrapeMetric(db *sql.DB, ch chan<- prometheus.Metric, metricDefinition Metric) error {
log.Debugln("Calling function ScrapeGenericValues()")
return ScrapeGenericValues(db, ch, metricDefinition.Context, metricDefinition.Labels,
metricDefinition.MetricsDesc, metricDefinition.MetricsType, metricDefinition.MetricsBuckets,
metricDefinition.FieldToAppend, metricDefinition.IgnoreZeroResult,
metricDefinition.Request)
}
// generic method for retrieving metrics.
func ScrapeGenericValues(db *sql.DB, ch chan<- prometheus.Metric, context string, labels []string,
metricsDesc map[string]string, metricsType map[string]string, metricsBuckets map[string]map[string]string, fieldToAppend string, ignoreZeroResult bool, request string) error {
metricsCount := 0
genericParser := func(row map[string]string) error {
// Construct labels value
labelsValues := []string{}
for _, label := range labels {
labelsValues = append(labelsValues, row[label])
}
// Construct Prometheus values to sent back
for metric, metricHelp := range metricsDesc {
value, err := strconv.ParseFloat(strings.TrimSpace(row[metric]), 64)
// If not a float, skip current metric
if err != nil {
log.Errorln("Unable to convert current value to float (metric=" + metric +
",metricHelp=" + metricHelp + ",value=<" + row[metric] + ">)")
continue
}
log.Debugln("Query result looks like: ", value)
Below is some example output of debug.
DEBU[0061] Query result looks like: 0 source="main.go:322"
DEBU[0061] Query result looks like: 0 source="main.go:322"
DEBU[0061] Query result looks like: 0 source="main.go:322"
DEBU[0061] Query result looks like: 0 source="main.go:322"
DEBU[0061] Query result looks like: -357 source="main.go:322"
DEBU[0061] Query result looks like: 0 source="main.go:322"
DEBU[0061] Query result looks like: 0 source="main.go:322"
DEBU[0061] Query result looks like: 0 source="main.go:322"
DEBU[0061] Query result looks like: 0 source="main.go:322"
DEBU[0061] Query result looks like: 0 source="main.go:322"
DEBU[0061] ScrapeGenericValues() - metricsCount: 10 source="main.go:397"
DEBU[0061] Successfully scraped metric: ohlprod map[qmxtm_genpicks_q11:Max time of queue entries on genpicks_q11 qmxtm_genpicks_q12:Max time of queue entries on genpicks_q12 qmxtm_genpicks_q13:Max time of queue entries on genpicks_q13 qmxtm_genpicks_q14:Max time of queue entries on genpicks_q14 qmxtm_genpicks_q15:Max time of queue entries on genpicks_q15 qmxtm_genpicks_q16:Max time of queue entries on genpicks_q16 qmxtm_genpicks_q17:Max time of queue entries on genpicks_q17 qmxtm_genpicks_q18:Max time of queue entries on genpicks_q18 qmxtm_genpicks_q19:Max time of queue entries on genpicks_q19 qmxtm_genpicks_q20:Max time of queue entries on genpicks_q20] 51.75887016s source="main.go:269"
I know from experience with this that the value of -357 means that a positive value of 3 was returned by the SQL query and then interpreted as -357 for some reason.
Other info: The working oracledb_exporter is running in our data center with os release shown below.
[someuser@somemachine ~]$ cat /etc/os-release
NAME="Oracle Linux Server"
VERSION="7.7"
ID="ol"
ID_LIKE="fedora"
VARIANT="Server"
VARIANT_ID="server"
VERSION_ID="7.7"
PRETTY_NAME="Oracle Linux Server 7.7"
ANSI_COLOR="0;31"
CPE_NAME="cpe:/o:oracle:linux:7:7:server"
HOME_URL="https://linux.oracle.com/"
BUG_REPORT_URL="https://bugzilla.oracle.com/"
ORACLE_BUGZILLA_PRODUCT="Oracle Linux 7"
ORACLE_BUGZILLA_PRODUCT_VERSION=7.7
ORACLE_SUPPORT_PRODUCT="Oracle Linux"
ORACLE_SUPPORT_PRODUCT_VERSION=7.7
[someuser@somemachine ~]$
The problem oracledb_exporter is running in Azure with os release shown below.
[someuser@somemachine oracledb_exporter]# cat /etc/os-release
NAME="Oracle Linux Server"
VERSION="8.4"
ID="ol"
ID_LIKE="fedora"
VARIANT="Server"
VARIANT_ID="server"
VERSION_ID="8.4"
PLATFORM_ID="platform:el8"
PRETTY_NAME="Oracle Linux Server 8.4"
ANSI_COLOR="0;31"
CPE_NAME="cpe:/o:oracle:linux:8:4:server"
HOME_URL="https://linux.oracle.com/"
BUG_REPORT_URL="https://bugzilla.oracle.com/"
ORACLE_BUGZILLA_PRODUCT="Oracle Linux 8"
ORACLE_BUGZILLA_PRODUCT_VERSION=8.4
ORACLE_SUPPORT_PRODUCT="Oracle Linux"
ORACLE_SUPPORT_PRODUCT_VERSION=8.4
[someuser@somemachine oracledb_exporter]#
Also, in my original post I referred to older version versus newer version of oracledb_exporter. However, I believe that it is the latest version of oracledb_exporter 0.3.2 on both my data center and my Azure machines.
Giving the above information as I try to see what the issue might be. Anyone with any suggestions to try or insight all assistance is appreciated. Thank you!
HansK-p
In a TOML file I have the below SQL query to pull max age of an Oracle queue entry for a given Oracle queue.
select (SELECT extract(day from 24*60*nvl(max(sysdate - from_tz(enq_time,'GMT') at TIME ZONE 'US/Central'),numtodsinterval(0,'minute'))) FROM alps.qt_rp_genpicks_q11) as qmxtm_genpicks_q11 from dual;The above query returns either 0 or a positive number representing the age in minutes of the oldest message on a queue. On a slightly older version of oracledb_exporter this returns the metric correctly, either 0 or a positive number.However, on the latest version of oracledb_exporter the query is running fine but the metric value returned is a strange negative behavior. If the SQL query value returned is 0 then the metric is 0. If the SQL query value returned is 1 then the metric value is -359. A SQL query value of 2 yields a metric value of -358 and so on.
Does anyone have any insight into this metric behavior?