17678: implement review feedback.

[arvados.git] / lib / costanalyzer / costanalyzer.go

diff --git a/lib/costanalyzer/costanalyzer.go b/lib/costanalyzer/costanalyzer.go
index 01a9d9fdb6b140280e89f63a30def7d77d2b365b..402e0ec81a160e677a1aefb8df82d589f01c3e20 100644 (file)
--- a/lib/costanalyzer/costanalyzer.go
+++ b/lib/costanalyzer/costanalyzer.go
@@ -35,6 +35,7 @@ type nodeInfo struct {
        // Modern
        ProviderType string
        Price        float64
        // Modern
        ProviderType string
        Price        float64

+       Preemptible  bool

 }
 
 type arrayFlags []string
 }
 
 type arrayFlags []string


@@ -56,12 +57,12 @@ func parseFlags(prog string, args []string, loader *config.Loader, logger *logru
        flags.Usage = func() {
                fmt.Fprintf(flags.Output(), `
 Usage:
        flags.Usage = func() {
                fmt.Fprintf(flags.Output(), `
 Usage:

-  %s [options ...]
+  %s [options ...] uuid [uuid ...]

 
        This program analyzes the cost of Arvados container requests. For each uuid
        supplied, it creates a CSV report that lists all the containers used to
        fulfill the container request, together with the machine type and cost of
 
        This program analyzes the cost of Arvados container requests. For each uuid
        supplied, it creates a CSV report that lists all the containers used to
        fulfill the container request, together with the machine type and cost of

-       each container.
+       each container. At least one uuid must be specified.

 
        When supplied with the uuid of a container request, it will calculate the
        cost of that container request and all its children.
 
        When supplied with the uuid of a container request, it will calculate the
        cost of that container request and all its children.


@@ -77,30 +78,42 @@ Usage:
        container was reused between several container requests, its cost will only
        be counted once.
 
        container was reused between several container requests, its cost will only
        be counted once.
 

-       To get the node costs, the progam queries the Arvados API for current cost
-       data for each node type used. This means that the reported cost always
-       reflects the cost data as currently defined in the Arvados API configuration
-       file.
-

        Caveats:
        Caveats:

-       - the Arvados API configuration cost data may be out of sync with the cloud
-       provider.
-       - when generating reports for older container requests, the cost data in the
-       Arvados API configuration file may have changed since the container request
-       was fulfilled. This program uses the cost data stored at the time of the
+
+       - This program uses the cost data from config.yml at the time of the

        execution of the container, stored in the 'node.json' file in its log
        execution of the container, stored in the 'node.json' file in its log

-       collection.
+       collection. If the cost data was not correctly configured at the time the
+       container was executed, the output from this program will be incorrect.
+
+       - If a container was run on a preemptible ("spot") instance, the cost data
+       reported by this program may be wildly inaccurate, because it does not have
+       access to the spot pricing in effect for the node then the container ran. The
+       UUID report file that is generated when the '-output' option is specified has
+       a column that indicates the preemptible state of the instance that ran the
+       container.
+
+       - This program does not take into account overhead costs like the time spent
+       starting and stopping compute nodes that run containers, the cost of the
+       permanent cloud nodes that provide the Arvados services, the cost of data
+       stored in Arvados, etc.
+
+       - When provided with a project uuid, subprojects will not be considered.

 
        In order to get the data for the uuids supplied, the ARVADOS_API_HOST and
        ARVADOS_API_TOKEN environment variables must be set.
 
 
        In order to get the data for the uuids supplied, the ARVADOS_API_HOST and
        ARVADOS_API_TOKEN environment variables must be set.
 

+       This program prints the total dollar amount from the aggregate cost
+       accounting across all provided uuids on stdout.
+
+       When the '-output' option is specified, a set of CSV files with cost details
+       will be written to the provided directory.
+

 Options:
 `, prog)
                flags.PrintDefaults()
        }
        loglevel := flags.String("log-level", "info", "logging `level` (debug, info, ...)")
 Options:
 `, prog)
                flags.PrintDefaults()
        }
        loglevel := flags.String("log-level", "info", "logging `level` (debug, info, ...)")

-       flags.StringVar(&resultsDir, "output", "", "output `directory` for the CSV reports (required)")
-       flags.Var(&uuids, "uuid", "object uuid. May be specified more than once. Also accepts a comma separated list of uuids (required)")
+       flags.StringVar(&resultsDir, "output", "", "output `directory` for the CSV reports")

        flags.BoolVar(&cache, "cache", true, "create and use a local disk cache of Arvados objects")
        err = flags.Parse(args)
        if err == flag.ErrHelp {
        flags.BoolVar(&cache, "cache", true, "create and use a local disk cache of Arvados objects")
        err = flags.Parse(args)
        if err == flag.ErrHelp {


@@ -111,17 +124,11 @@ Options:
                exitCode = 2
                return
        }
                exitCode = 2
                return
        }

+       uuids = flags.Args()

 
        if len(uuids) < 1 {
                flags.Usage()
 
        if len(uuids) < 1 {
                flags.Usage()

-               err = fmt.Errorf("Error: no uuid(s) provided")
-               exitCode = 2
-               return
-       }
-
-       if resultsDir == "" {
-               flags.Usage()
-               err = fmt.Errorf("Error: output directory must be specified")
+               err = fmt.Errorf("error: no uuid(s) provided")

                exitCode = 2
                return
        }
                exitCode = 2
                return
        }


@@ -182,7 +189,7 @@ func addContainerLine(logger *logrus.Logger, node nodeInfo, cr arvados.Container
                size = node.ProviderType
        }
        cost = delta.Seconds() / 3600 * price
                size = node.ProviderType
        }
        cost = delta.Seconds() / 3600 * price

-       csv += size + "," + strconv.FormatFloat(price, 'f', 8, 64) + "," + strconv.FormatFloat(cost, 'f', 8, 64) + "\n"
+       csv += size + "," + fmt.Sprintf("%+v", node.Preemptible) + "," + strconv.FormatFloat(price, 'f', 8, 64) + "," + strconv.FormatFloat(cost, 'f', 8, 64) + "\n"

        return
 }
 
        return
 }
 


@@ -286,7 +293,7 @@ func loadObject(logger *logrus.Logger, ac *arvados.Client, path string, uuid str
 
 func getNode(arv *arvadosclient.ArvadosClient, ac *arvados.Client, kc *keepclient.KeepClient, cr arvados.ContainerRequest) (node nodeInfo, err error) {
        if cr.LogUUID == "" {
 
 func getNode(arv *arvadosclient.ArvadosClient, ac *arvados.Client, kc *keepclient.KeepClient, cr arvados.ContainerRequest) (node nodeInfo, err error) {
        if cr.LogUUID == "" {

-               err = errors.New("No log collection")
+               err = errors.New("no log collection")

                return
        }
 
                return
        }
 


@@ -370,7 +377,7 @@ func generateCrCsv(logger *logrus.Logger, uuid string, arv *arvadosclient.Arvado
 
        cost = make(map[string]float64)
 
 
        cost = make(map[string]float64)
 

-       csv := "CR UUID,CR name,Container UUID,State,Started At,Finished At,Duration in seconds,Compute node type,Hourly node cost,Total cost\n"
+       csv := "CR UUID,CR name,Container UUID,State,Started At,Finished At,Duration in seconds,Compute node type,Preemptible,Hourly node cost,Total cost\n"

        var tmpCsv string
        var tmpTotalCost float64
        var totalCost float64
        var tmpCsv string
        var tmpTotalCost float64
        var totalCost float64


@@ -387,7 +394,10 @@ func generateCrCsv(logger *logrus.Logger, uuid string, arv *arvadosclient.Arvado
                if !ok {
                        return nil, fmt.Errorf("error: collection %s does not have a 'container_request' property", uuid)
                }
                if !ok {
                        return nil, fmt.Errorf("error: collection %s does not have a 'container_request' property", uuid)
                }

-               crUUID = value.(string)
+               crUUID, ok = value.(string)
+               if !ok {
+                       return nil, fmt.Errorf("error: collection %s does not have a 'container_request' property of the string type", uuid)
+               }

        }
 
        // This is a container request, find the container
        }
 
        // This is a container request, find the container


@@ -397,7 +407,7 @@ func generateCrCsv(logger *logrus.Logger, uuid string, arv *arvadosclient.Arvado
                return nil, fmt.Errorf("error loading cr object %s: %s", uuid, err)
        }
        var container arvados.Container
                return nil, fmt.Errorf("error loading cr object %s: %s", uuid, err)
        }
        var container arvados.Container

-       err = loadObject(logger, ac, uuid, cr.ContainerUUID, cache, &container)
+       err = loadObject(logger, ac, crUUID, cr.ContainerUUID, cache, &container)

        if err != nil {
                return nil, fmt.Errorf("error loading container object %s: %s", cr.ContainerUUID, err)
        }
        if err != nil {
                return nil, fmt.Errorf("error loading container object %s: %s", cr.ContainerUUID, err)
        }


@@ -426,7 +436,7 @@ func generateCrCsv(logger *logrus.Logger, uuid string, arv *arvadosclient.Arvado
        if err != nil {
                return nil, fmt.Errorf("error querying container_requests: %s", err.Error())
        }
        if err != nil {
                return nil, fmt.Errorf("error querying container_requests: %s", err.Error())
        }

-       logger.Infof("Collecting child containers for container request %s", uuid)
+       logger.Infof("Collecting child containers for container request %s", crUUID)

        for _, cr2 := range childCrs.Items {
                logger.Info(".")
                node, err := getNode(arv, ac, kc, cr2)
        for _, cr2 := range childCrs.Items {
                logger.Info(".")
                node, err := getNode(arv, ac, kc, cr2)


@@ -435,7 +445,7 @@ func generateCrCsv(logger *logrus.Logger, uuid string, arv *arvadosclient.Arvado
                }
                logger.Debug("\nChild container: " + cr2.ContainerUUID + "\n")
                var c2 arvados.Container
                }
                logger.Debug("\nChild container: " + cr2.ContainerUUID + "\n")
                var c2 arvados.Container

-               err = loadObject(logger, ac, uuid, cr2.ContainerUUID, cache, &c2)
+               err = loadObject(logger, ac, cr.UUID, cr2.ContainerUUID, cache, &c2)

                if err != nil {
                        return nil, fmt.Errorf("error loading object %s: %s", cr2.ContainerUUID, err)
                }
                if err != nil {
                        return nil, fmt.Errorf("error loading object %s: %s", cr2.ContainerUUID, err)
                }


@@ -448,13 +458,15 @@ func generateCrCsv(logger *logrus.Logger, uuid string, arv *arvadosclient.Arvado
 
        csv += "TOTAL,,,,,,,,," + strconv.FormatFloat(totalCost, 'f', 8, 64) + "\n"
 
 
        csv += "TOTAL,,,,,,,,," + strconv.FormatFloat(totalCost, 'f', 8, 64) + "\n"
 

-       // Write the resulting CSV file
-       fName := resultsDir + "/" + uuid + ".csv"
-       err = ioutil.WriteFile(fName, []byte(csv), 0644)
-       if err != nil {
-               return nil, fmt.Errorf("error writing file with path %s: %s", fName, err.Error())
+       if resultsDir != "" {
+               // Write the resulting CSV file
+               fName := resultsDir + "/" + crUUID + ".csv"
+               err = ioutil.WriteFile(fName, []byte(csv), 0644)
+               if err != nil {
+                       return nil, fmt.Errorf("error writing file with path %s: %s", fName, err.Error())
+               }
+               logger.Infof("\nUUID report in %s\n\n", fName)

        }
        }

-       logger.Infof("\nUUID report in %s\n\n", fName)

 
        return
 }
 
        return
 }


@@ -464,10 +476,12 @@ func costanalyzer(prog string, args []string, loader *config.Loader, logger *log
        if exitcode != 0 {
                return
        }
        if exitcode != 0 {
                return
        }

-       err = ensureDirectory(logger, resultsDir)
-       if err != nil {
-               exitcode = 3
-               return
+       if resultsDir != "" {
+               err = ensureDirectory(logger, resultsDir)
+               if err != nil {
+                       exitcode = 3
+                       return
+               }

        }
 
        // Arvados Client setup
        }
 
        // Arvados Client setup


@@ -503,7 +517,7 @@ func costanalyzer(prog string, args []string, loader *config.Loader, logger *log
                        var crCsv map[string]float64
                        crCsv, err = generateCrCsv(logger, uuid, arv, ac, kc, resultsDir, cache)
                        if err != nil {
                        var crCsv map[string]float64
                        crCsv, err = generateCrCsv(logger, uuid, arv, ac, kc, resultsDir, cache)
                        if err != nil {

-                               err = fmt.Errorf("Error generating CSV for uuid %s: %s", uuid, err.Error())
+                               err = fmt.Errorf("error generating CSV for uuid %s: %s", uuid, err.Error())

                                exitcode = 2
                                return
                        }
                                exitcode = 2
                                return
                        }


@@ -515,7 +529,11 @@ func costanalyzer(prog string, args []string, loader *config.Loader, logger *log
                        // It is identified by the user uuid. As such, cost analysis for the
                        // "Home" project is not supported by this program. Skip this uuid, but
                        // keep going.
                        // It is identified by the user uuid. As such, cost analysis for the
                        // "Home" project is not supported by this program. Skip this uuid, but
                        // keep going.

-                       logger.Errorf("Cost analysis is not supported for the 'Home' project: %s", uuid)
+                       logger.Errorf("cost analysis is not supported for the 'Home' project: %s", uuid)
+               } else {
+                       logger.Errorf("this argument does not look like a uuid: %s\n", uuid)
+                       exitcode = 3
+                       return

                }
        }
 
                }
        }
 


@@ -539,14 +557,20 @@ func costanalyzer(prog string, args []string, loader *config.Loader, logger *log
 
        csv += "TOTAL," + strconv.FormatFloat(total, 'f', 8, 64) + "\n"
 
 
        csv += "TOTAL," + strconv.FormatFloat(total, 'f', 8, 64) + "\n"
 

-       // Write the resulting CSV file
-       aFile := resultsDir + "/" + time.Now().Format("2006-01-02-15-04-05") + "-aggregate-costaccounting.csv"
-       err = ioutil.WriteFile(aFile, []byte(csv), 0644)
-       if err != nil {
-               err = fmt.Errorf("Error writing file with path %s: %s", aFile, err.Error())
-               exitcode = 1
-               return
+       if resultsDir != "" {
+               // Write the resulting CSV file
+               aFile := resultsDir + "/" + time.Now().Format("2006-01-02-15-04-05") + "-aggregate-costaccounting.csv"
+               err = ioutil.WriteFile(aFile, []byte(csv), 0644)
+               if err != nil {
+                       err = fmt.Errorf("error writing file with path %s: %s", aFile, err.Error())
+                       exitcode = 1
+                       return
+               }
+               logger.Infof("Aggregate cost accounting for all supplied uuids in %s\n", aFile)

        }
        }

-       logger.Infof("Aggregate cost accounting for all supplied uuids in %s\n", aFile)
+
+       // Output the total dollar amount on stdout
+       fmt.Fprintf(stdout, "%s\n", strconv.FormatFloat(total, 'f', 8, 64))
+

        return
 }
        return
 }