Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Corrected how the --partial_dir_finish flag needs to be used. #96

Merged
merged 7 commits into from
Oct 7, 2024
Merged

Corrected how the --partial_dir_finish flag needs to be used. #96

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
7 commits
Select commit Hold shift + click to select a range
3339de2
Corrected how the --partial_dir_finish flag needs to be used.
mjkw31 Oct 1, 2024
d52d512
Delint'd.
mjkw31 Oct 1, 2024
d56a164
Add additional help text for the --partial_dir_finish flag.
mjkw31 Oct 2, 2024
6fa0b9a
Correct var name.
mjkw31 Oct 2, 2024
835eec4
Delint'd.
mjkw31 Oct 2, 2024
164c546
Remove path requirement when using --partial_dir_finish.
mjkw31 Oct 2, 2024
0eca1e9
Add end-to-end tests using --partial_dir_finish flag.
mjkw31 Oct 2, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 2 additions & 2 deletions cmd/cron.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -108,8 +108,8 @@ func init() {
"from specified directory after merging")
cronCmd.Flags().BoolVarP(&createPartial, "create_partial_dir", "p", false, "perform the walk, "+
"stat, and combine steps only")
cronCmd.Flags().BoolVarP(&finishPartial, "partial_dir_finish", "z", false, "perform the basedir "+
"and tidy step on a partial run")
cronCmd.Flags().StringVarP(&finishPartial, "partial_dir_finish", "z", "", "perform the basedir "+
"and tidy step on a partial run, requires the name of the unique subdirectory the partial run files are in")
cronCmd.Flags().IntVarP(&multiInodes, "inodes_per_stat", "n",
defaultInodesPerJob, "number of inodes per parallel stat job")
cronCmd.Flags().IntVarP(&multiStatJobs, "num_stat_jobs", "j",
Expand Down
20 changes: 14 additions & 6 deletions cmd/multi.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -56,7 +56,7 @@ var (
partialDirMerge string
partialDirClean bool
createPartial bool
finishPartial bool
finishPartial string
multiInodes int
multiStatJobs int
multiCh string
Expand Down Expand Up @@ -163,8 +163,8 @@ func init() {
"from specified directory after merging")
multiCmd.Flags().BoolVarP(&createPartial, "create_partial_dir", "p", false, "perform the walk, "+
"stat, and combine steps only")
multiCmd.Flags().BoolVarP(&finishPartial, "partial_dir_finish", "z", false, "perform the basedir "+
"and tidy step on a partial run")
multiCmd.Flags().StringVarP(&finishPartial, "partial_dir_finish", "z", "", "perform the basedir "+
"and tidy step on a partial run, requires the name of the unique subdirectory the partial run files are in")
multiCmd.Flags().IntVarP(&multiInodes, "inodes_per_stat", "n",
defaultInodesPerJob, "number of inodes per parallel stat job")
multiCmd.Flags().IntVarP(&multiStatJobs, "num_stat_jobs", "j",
Expand All @@ -189,7 +189,7 @@ func checkMultiArgs(args []string) {
checkStandardFlags()
}

if len(args) == 0 {
if len(args) == 0 && finishPartial == "" {
die("at least 1 directory of interest must be supplied")
}
}
Expand All @@ -213,15 +213,15 @@ func doMultiScheduling(args []string, workDir, forcedQueue, queuesToAvoid string
s, d := newScheduler(workDir, forcedQueue, queuesToAvoid, sudo)
defer d()

unique := scheduler.UniqueString()
unique := uniqueOrPartial(finishPartial)
outputRoot := filepath.Join(workDir, unique)

err := os.MkdirAll(outputRoot, userGroupPerm)
if err != nil {
return err
}

if !finishPartial { //nolint:nestif
if finishPartial == "" { //nolint:nestif
scheduleWalkJobs(outputRoot, args, unique, multiStatJobs, multiInodes, multiCh, forcedQueue, queuesToAvoid, s)

if partialDirMerge != "" {
Expand All @@ -242,6 +242,14 @@ func doMultiScheduling(args []string, workDir, forcedQueue, queuesToAvoid string
return nil
}

func uniqueOrPartial(partial string) string {
if partial == "" {
return scheduler.UniqueString()
}

return partial
}

// scheduleWalkJobs adds a 'wrstat walk' job to wr's queue for each desired
// path. The second scheduler is used to add combine jobs, which need a memory
// override.
Expand Down
45 changes: 40 additions & 5 deletions main_test.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -537,12 +537,11 @@ func testPartial(t *testing.T, sudo bool, subcommand []string, date string, walk
Convey("…finishing the partial run runs the correct jobs", func() {
workingDir := t.TempDir()
_, _, jobs, err := runWRStat(append(subcommand, "-w", workingDir, "-f", "final_output", "-q", "quota_file",
"-o", "owners_file", "-z", "/some/path", "/some-other/path")...)
"-o", "owners_file", "-z", repGroup, "/some/path", "/some-other/path")...)
So(err, ShouldBeNil)

So(len(jobs), ShouldEqual, 2)

repGroup := jobs[0].RepGroup[len(jobs[0].RepGroup)-20:]
expectation := []*jobqueue.Job{
{
Cmd: fmt.Sprintf("%s basedir -q \"quota_file\" -o \"owners_file\" \"%s/%s\" \"final_output\"",
Expand Down Expand Up @@ -1650,6 +1649,7 @@ waitForJobs() {
mkdir -p /tmp/working/partial/;
mkdir -p /tmp/working/complete/;
mkdir -p /tmp/final/;
mkdir -p /tmp/final/simple/;

yes y | WR_RunnerExecShell=sh wr manager start -s local --max_ram -1 --max_cores -1;

Expand All @@ -1660,6 +1660,10 @@ wrstat multi -m 0 -w /tmp/working/complete/ -f /tmp/final/ -l `+
`/tmp/working/partial -q /tmp/quota -b /tmp/basedirs -o /tmp/owners /objects/*;
waitForJobs;

wrstat multi -m 0 -z "$(basename /tmp/working/partial/*)" -w /tmp/working/partial/ `+
`-f /tmp/final/simple -q /tmp/quota -b /tmp/basedirs -o /tmp/owners;
waitForJobs;

stop;`)
writeFileString(t, filepath.Join(tmpTemp, "owners"), "")
writeFileString(t, filepath.Join(tmpTemp, "quota"), "")
Expand Down Expand Up @@ -1784,7 +1788,7 @@ stop;`)
"U%[2]d\t\t/objects/store2/part0/teams/team2\t%[6]d\t1000\t0\t1\t0\tOK\n"+
"U%[3]d\t\t/objects/store1/data/temp\t%[6]d\t6000\t0\t3\t0\tOK\n"+
"U%[4]d\t\t/objects/store2/part1/other/my\nDir\t%[6]d\t2048\t0\t1\t0\tOK\n"+
"U%[5]d\t\t/objects/store2/part1/other/my\nDir\t19997\t2048\t0\t1\t0\tOK\n"+
"U%[5]d\t\t/objects/store2/part1/other/my\nDir\t%[6]d\t2048\t0\t1\t0\tOK\n"+
"U%[5]d\t\t/simple/E\t%[6]d\t2\t0\t1\t0\tOK",
UserA, UserB, UserC, UserD, UserE,
time.Now().Unix()/86400),
Expand Down Expand Up @@ -1953,6 +1957,37 @@ stop;`)
encode.Base64Encode("/objects/store3/aFile")+"\t512\t%d\t%d\t160\t160\t160\tf\t\x00\t1\t34\n"+
encode.Base64Encode("/objects/store3")+"\t0\t0\t0\t160\t160\t160\td\t\x00\t2\t32",
UserA, GroupA),

"simple/*basedirs.userusage.tsv": fmt.Sprintf(``+
"U%[1]d\t\t/simple/A\t%[3]d\t1\t0\t1\t0\tOK\n"+
"U%[2]d\t\t/simple/E\t%[3]d\t2\t0\t1\t0\tOK",
UserA, UserE,
time.Now().Unix()/86400),
"simple/*basedirs.groupusage.tsv": fmt.Sprintf(``+
"G%[1]d\t\t/simple/A\t%[3]d\t1\t0\t1\t0\tNot OK\n"+
"G%[2]d\t\t/simple/E\t%[3]d\t2\t0\t1\t0\tNot OK",
GroupA, GroupE,
time.Now().Unix()/86400),
"simple/????????_A.*.bygroup": fmt.Sprintf("G%d\tU%d\t1\t1", GroupA, UserA),
"simple/????????_E.*.bygroup": fmt.Sprintf("G%d\tU%d\t1\t2", GroupE, UserE),
"simple/????????_A.*.byusergroup.gz": fmt.Sprintf(``+
"U%[1]d\tG%[2]d\t"+encode.Base64Encode("/")+"\t1\t1\n"+
"U%[1]d\tG%[2]d\t"+encode.Base64Encode("/simple")+"\t1\t1\n"+
"U%[1]d\tG%[2]d\t"+encode.Base64Encode("/simple/A")+"\t1\t1\n", UserA, GroupA),
"simple/????????_E.*.byusergroup.gz": fmt.Sprintf(``+
"U%[1]d\tG%[2]d\t"+encode.Base64Encode("/")+"\t1\t2\n"+
"U%[1]d\tG%[2]d\t"+encode.Base64Encode("/simple")+"\t1\t2\n"+
"U%[1]d\tG%[2]d\t"+encode.Base64Encode("/simple/E")+"\t1\t2\n", UserE, GroupE),
"simple/????????_A.*.logs.gz": "",
"simple/????????_E.*.logs.gz": "",
"simple/????????_A.*.stats.gz": fmt.Sprintf(""+
encode.Base64Encode("/simple/A/a.file")+"\t1\t%[1]d\t%[2]d\t166\t166\t166\tf\t\x00\t1\t34\n"+
encode.Base64Encode("/simple/A")+"\t0\t%[1]d\t%[2]d\t166\t166\t166\td\t\x00\t2\t32",
UserA, GroupA),
"simple/????????_E.*.stats.gz": fmt.Sprintf(""+
encode.Base64Encode("/simple/E/b.tmp")+"\t2\t%[1]d\t%[2]d\t171\t171\t171\tf\t\x00\t2\t34\n"+
encode.Base64Encode("/simple/E")+"\t0\t%[1]d\t%[2]d\t171\t171\t171\td\t\x00\t3\t32",
UserE, GroupE),
} {
files, errr := fs.Glob(os.DirFS(tmpTemp), filepath.Join("final", file))
So(errr, ShouldBeNil)
Expand All @@ -1977,7 +2012,7 @@ stop;`)
"%[2]d\t\t/objects/store2/part0/teams/team2\t%[6]d\t1000\t0\t1\t0\tOK\n"+
"%[3]d\t\t/objects/store1/data/temp\t%[6]d\t6000\t0\t3\t0\tOK\n"+
"%[4]d\t\t/objects/store2/part1/other/my\nDir\t%[6]d\t2048\t0\t1\t0\tOK\n"+
"%[5]d\t\t/objects/store2/part1/other/my\nDir\t19997\t2048\t0\t1\t0\tOK\n"+
"%[5]d\t\t/objects/store2/part1/other/my\nDir\t%[6]d\t2048\t0\t1\t0\tOK\n"+
"%[5]d\t\t/simple/E\t%[6]d\t2\t0\t1\t0\tOK\n",
UserA, UserB, UserC, UserD, UserE,
time.Now().Unix()/86400)
Expand All @@ -1991,7 +2026,7 @@ stop;`)
"%[1]d\t\t/simple/A\t%[5]d\t1\t0\t1\t0\tNot OK\n"+
"%[2]d\t\t/objects/store2/part0/teams/team1\t%[5]d\t200\t0\t1\t0\tNot OK\n"+
"%[2]d\t\t/objects/store2/part0/teams/team2\t%[5]d\t1000\t0\t1\t0\tNot OK\n"+
"%[2]d\t\t/objects/store2/part1/other/my\nDir\t19997\t2048\t0\t1\t0\tNot OK\n"+
"%[2]d\t\t/objects/store2/part1/other/my\nDir\t%[5]d\t2048\t0\t1\t0\tNot OK\n"+
"%[3]d\t\t/objects/store2/important/docs\t\t%[5]d\t1200\t0\t1\t0\tNot OK\n"+
"%[4]d\t\t/simple/E\t%[5]d\t2\t0\t1\t0\tNot OK\n",
GroupA, GroupB, GroupD, GroupE,
Expand Down
Loading