Invalid periodic job cron syntax causes crash loop

[sevagh]

Nomad version

Nomad v0.7.1 (0b295d399d00199cfab4621566babd25987ba06e)

Operating system and Environment details

Nomad cluster running across a handful of Linux hosts (kernel 3.6, Debian 8).

Issue

When a periodic job with invalid syntax is submitted, e.g.:

"Spec": "1  15-0  *  *  1-5",
"SpecType": "cron"

The underlying cron library (gorhill/cronexpr#35) throws an Index out of range panic. It's the expected behavior by the library (and confirmed by the maintainer), 15-0 is incorrect cron syntax and the library doesn't guarantee behavior for invalid syntax.

However, this makes Nomad get the panic when trying to schedule the job, and they all went into a crash loop:

func (p *PeriodicConfig) Next(fromTime time.Time) time.Time {
	switch p.SpecType {
	case PeriodicSpecCron:
		if e, err := cronexpr.Parse(p.Spec); err == nil {
			return e.Next(fromTime)
		}

https://github.com/hashicorp/nomad/blob/master/nomad/structs/structs.go#L2660

Not sure what the steps are going forward - if there's Nomad-side validation of cron syntax, then why rely on a third-party cron library? Is panic-catching considered idiomatic Go?

Reproduction steps

Pass a periodic job with invalid cron syntax.

Nomad Server logs (if appropriate)

Apr 26 09:28:53 baratheon nomad[612]: panic: runtime error: index out of range
Apr 26 09:28:53 baratheon nomad[612]: goroutine 495 [running]:
Apr 26 09:28:53 baratheon nomad[612]: github.com/hashicorp/nomad/vendor/github.com/gorhill/cronexpr.(*Expression).nextDayOfMonth(0xc423e702a0, 0x3268b4de, 0x
ed273f545, 0x0, 0x0, 0x15, 0x5)
Apr 26 09:28:53 baratheon nomad[612]: /opt/gopath/src/github.com/hashicorp/nomad/vendor/github.com/gorhill/cronexpr/cronexpr_next.go:115 +0x254
Apr 26 09:28:53 baratheon nomad[612]: github.com/hashicorp/nomad/vendor/github.com/gorhill/cronexpr.(*Expression).Next(0xc423e702a0, 0x3268b4de, 0xed273f545,
 0x0, 0x0, 0x1f0e780, 0x362ef45e)
Apr 26 09:28:53 baratheon nomad[612]: /opt/gopath/src/github.com/hashicorp/nomad/vendor/github.com/gorhill/cronexpr/cronexpr.go:210 +0x699
Apr 26 09:28:53 baratheon nomad[612]: github.com/hashicorp/nomad/nomad/structs.(*PeriodicConfig).Next(0xc4234800a0, 0x3268b4de, 0xed273f545, 0x0, 0x3268b4de,
 0xed273f545, 0x0)
Apr 26 09:28:53 baratheon nomad[612]: /opt/gopath/src/github.com/hashicorp/nomad/nomad/structs/structs.go:2353 +0x3a8
Apr 26 09:28:53 baratheon nomad[612]: github.com/hashicorp/nomad/nomad.(*PeriodicDispatch).Add(0xc420492780, 0xc423656340, 0x0, 0x0)
Apr 26 09:28:53 baratheon nomad[612]: /opt/gopath/src/github.com/hashicorp/nomad/nomad/periodic.go:224 +0x2c8
Apr 26 09:28:53 baratheon nomad[612]: github.com/hashicorp/nomad/nomad.(*Server).restorePeriodicDispatcher(0xc42018a1e0, 0x1, 0x0)
Apr 26 09:28:53 baratheon nomad[612]: /opt/gopath/src/github.com/hashicorp/nomad/nomad/leader.go:363 +0x1cf
Apr 26 09:28:53 baratheon nomad[612]: github.com/hashicorp/nomad/nomad.(*Server).establishLeadership(0xc42018a1e0, 0xc42256cae0, 0x3, 0xbeb09d71679c157c)
Apr 26 09:28:53 baratheon nomad[612]: /opt/gopath/src/github.com/hashicorp/nomad/nomad/leader.go:204 +0x24d
Apr 26 09:28:53 baratheon nomad[612]: github.com/hashicorp/nomad/nomad.(*Server).leaderLoop(0xc42018a1e0, 0xc42256cae0)
Apr 26 09:28:53 baratheon nomad[612]: /opt/gopath/src/github.com/hashicorp/nomad/nomad/leader.go:105 +0x552
Apr 26 09:28:53 baratheon nomad[612]: github.com/hashicorp/nomad/nomad.(*Server).monitorLeadership.func1(0xc4257d6000, 0xc42018a1e0, 0xc42256cae0)
Apr 26 09:28:53 baratheon nomad[612]: /opt/gopath/src/github.com/hashicorp/nomad/nomad/leader.go:60 +0x5b
Apr 26 09:28:53 baratheon nomad[612]: created by github.com/hashicorp/nomad/nomad.(*Server).monitorLeadership
Apr 26 09:28:53 baratheon nomad[612]: /opt/gopath/src/github.com/hashicorp/nomad/nomad/leader.go:58 +0x1f9

[github-actions]

I'm going to lock this issue because it has been closed for 120 days ⏳. This helps our maintainers find and focus on the active issues.
If you have found a problem that seems similar to this, please open a new issue and complete the issue template so we can capture all the details necessary to investigate further.