package keepclient
import (
- "git.curoverse.com/arvados.git/sdk/go/streamer"
+ "crypto/md5"
"errors"
"fmt"
+ "git.curoverse.com/arvados.git/sdk/go/streamer"
"io"
"io/ioutil"
"log"
"net/http"
"os"
- "strconv"
"strings"
+ "time"
)
type keepDisk struct {
+ Uuid string `json:"uuid"`
Hostname string `json:"service_host"`
Port int `json:"service_port"`
SSL bool `json:"service_ssl_flag"`
SvcType string `json:"service_type"`
}
+func Md5String(s string) string {
+ return fmt.Sprintf("%x", md5.Sum([]byte(s)))
+}
+
func (this *KeepClient) DiscoverKeepServers() error {
if prx := os.Getenv("ARVADOS_KEEP_PROXY"); prx != "" {
- this.SetServiceRoots([]string{prx})
+ sr := map[string]string{"proxy": prx}
+ this.SetServiceRoots(sr)
this.Using_proxy = true
return nil
}
}
listed := make(map[string]bool)
- service_roots := make([]string, 0, len(m.Items))
+ service_roots := make(map[string]string)
for _, element := range m.Items {
n := ""
// Skip duplicates
if !listed[url] {
listed[url] = true
- service_roots = append(service_roots, url)
+ service_roots[element.Uuid] = url
}
if element.SvcType == "proxy" {
this.Using_proxy = true
return nil
}
-func (this KeepClient) shuffledServiceRoots(hash string) (pseq []string) {
- // Build an ordering with which to query the Keep servers based on the
- // contents of the hash. "hash" is a hex-encoded number at least 8
- // digits (32 bits) long
-
- // seed used to calculate the next keep server from 'pool' to be added
- // to 'pseq'
- seed := hash
-
- // Keep servers still to be added to the ordering
- service_roots := this.ServiceRoots()
- pool := make([]string, len(service_roots))
- copy(pool, service_roots)
-
- // output probe sequence
- pseq = make([]string, 0, len(service_roots))
-
- // iterate while there are servers left to be assigned
- for len(pool) > 0 {
-
- if len(seed) < 8 {
- // ran out of digits in the seed
- if len(pseq) < (len(hash) / 4) {
- // the number of servers added to the probe
- // sequence is less than the number of 4-digit
- // slices in 'hash' so refill the seed with the
- // last 4 digits.
- seed = hash[len(hash)-4:]
- }
- seed += hash
- }
-
- // Take the next 8 digits (32 bytes) and interpret as an integer,
- // then modulus with the size of the remaining pool to get the next
- // selected server.
- probe, _ := strconv.ParseUint(seed[0:8], 16, 32)
- probe %= uint64(len(pool))
-
- // Append the selected server to the probe sequence and remove it
- // from the pool.
- pseq = append(pseq, pool[probe])
- pool = append(pool[:probe], pool[probe+1:]...)
-
- // Remove the digits just used from the seed
- seed = seed[8:]
- }
- return pseq
-}
-
type uploadStatus struct {
err error
url string
}
func (this KeepClient) uploadToKeepServer(host string, hash string, body io.ReadCloser,
- upload_status chan<- uploadStatus, expectedLength int64) {
+ upload_status chan<- uploadStatus, expectedLength int64, tag string) {
- log.Printf("Uploading %s to %s", hash, host)
+ log.Printf("[%v] Begin upload %s to %s", tag, hash, host)
var req *http.Request
var err error
if resp.StatusCode == http.StatusOK {
upload_status <- uploadStatus{nil, url, resp.StatusCode, rep, locator}
} else {
- upload_status <- uploadStatus{errors.New(resp.Status), url, resp.StatusCode, rep, locator}
+ upload_status <- uploadStatus{errors.New(resp.Status), url, resp.StatusCode, rep, string(respbody)}
}
}
tr *streamer.AsyncStream,
expectedLength int64) (locator string, replicas int, err error) {
+ // Take the hash of locator and timestamp in order to identify this
+ // specific transaction in log statements.
+ tag := fmt.Sprintf("%x", md5.Sum([]byte(locator+time.Now().String())))[0:8]
+
// Calculate the ordering for uploading to servers
- sv := this.shuffledServiceRoots(hash)
+ sv := NewRootSorter(this.ServiceRoots(), hash).GetSortedRoots()
// The next server to try contacting
next_server := 0
defer close(upload_status)
// Desired number of replicas
-
remaining_replicas := this.Want_replicas
for remaining_replicas > 0 {
for active < remaining_replicas {
// Start some upload requests
if next_server < len(sv) {
- go this.uploadToKeepServer(sv[next_server], hash, tr.MakeStreamReader(), upload_status, expectedLength)
+ go this.uploadToKeepServer(sv[next_server], hash, tr.MakeStreamReader(), upload_status, expectedLength, tag)
next_server += 1
active += 1
} else {
// Now wait for something to happen.
status := <-upload_status
+ log.Printf("[%v] Upload to %v status code: %v remaining replicas: %v active: %v",
+ tag, status.url, status.statusCode, remaining_replicas, active)
if status.statusCode == 200 {
// good news!
remaining_replicas -= status.replicas_stored
locator = status.response
} else {
// writing to keep server failed for some reason
- log.Printf("Keep server put to %v failed with '%v'",
- status.url, status.err)
+ log.Printf("[%v] Upload to %v failed with error '%v', response '%v'",
+ tag, status.url, status.statusCode, status.err, status.response)
}
active -= 1
- log.Printf("Upload to %v status code: %v remaining replicas: %v active: %v", status.url, status.statusCode, remaining_replicas, active)
+
}
return locator, this.Want_replicas, nil