More details in state description

[msvoelker]

In general, I think we should add more details in the state description, since this is the main part of the text. That's why I herewith "reopened" the issue #5.

Here is the current last sentence of the SEARCHING state description.

The state is exited when the PROBE_COUNT reaches MAX_PROBES, a received PTB message is validated, a probe of size MAX_PMTU is acknowledged, or a black hole is detected.

Here is what I would write instead.

The SEARCHING concludes and the state transitions to SEARCH_COMPLETE, when a probe fails (PROBE_COUNT reaches MAX_PROBES), or when a validated PTB message reports a MTU that equals the last successful probed size (PTB_SIZE = PLPMTU). When a previously successful probed packet size is no longer supported by the path (black hole is detected), the state transitions to ERROR.

What do you think?
@adventureloop @gorryfair

[gorryfair]

On 14/06/2019, 16:08, msvoelker wrote: In general, I think we should add more details in the state description, since this is the main part of the text. That's why I herewith "reopened" the issue #5 <#5>. Here is the current last sentence of the SEARCHING state description. The state is exited when the PROBE_COUNT reaches MAX_PROBES, a received PTB message is validated, a probe of size MAX_PMTU is acknowledged, or a black hole is detected. Here is what I would write instead. The SEARCHING concludes and the state transitions to SEARCH_COMPLETE, when a probe fails (PROBE_COUNT reaches MAX_PROBES),

The above seems correct.

or when a validated PTB message reports a MTU that equals the last successful probed size (PTB_SIZE = PLPMTU).

It's still possible to probe higher than the PTB_SIZE... It may not work, but I don't see why the searrch absolutely has to exit, that's a design choice - but it would be OK to allow the search to terminate under that condition.

When a previously successful probed packet size is no longer supported by the path (black hole is detected), the state transitions to ERROR.

That's not true - Error is entered when no size is supported within the allowed range; or where the results are so inconsistent that it is not possible to determine reliably the actual PMTU.

What do you think? @adventureloop <https://github.com/adventureloop> @gorryfair <https://github.com/gorryfair> — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#9?email_source=notifications&email_token=ABYLLEVTVFDZG7VZELTWDJDP2OX77A5CNFSM4HYIX2X2YY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4GZSRTXA>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ABYLLEQJEHTJMBUGELYE5ZLP2OX77ANCNFSM4HYIX2XQ>.

Gorry

[msvoelker]

When a previously successful probed packet size is no longer supported by the path (black hole is detected), the state transitions to ERROR.

That's not true - Error is entered when no size is supported within the allowed range; or where the results are so inconsistent that it is not possible to determine reliably the actual PMTU.

Oh, you are right. My fault. The state diagram tells us to move to BASE then. So, I would write.

When a previously successful probed packet size is no longer supported by the path (black hole is detected), the state transitions to BASE.

[msvoelker]

or when a validated PTB message reports a MTU that equals the last successful probed size (PTB_SIZE = PLPMTU).

It's still possible to probe higher than the PTB_SIZE... It may not work, but I don't see why the searrch absolutely has to exit, that's a design choice - but it would be OK to allow the search to terminate under that condition.

Have you had this case in mind?
Probe S0 --> success
Probe S1 > S0 --> PTB with PTB_SIZE = S0
Probe S2, with S0 = PTB_SIZE < S2 < S1 --> may work

If yes, note issue #7 which describes this independently of a PTB.
If no, I'm not sure what you mean.

[tuexen]

All information being in the diagram should be in the text.

[msvoelker]

All information being in the diagram should be in the text.

Great! First the diagram must match to the two definitions for Black Hole #8. After that, I will update the text.

[msvoelker]

Since #8 is closed now, I looked into our state diagram description again. I found a few paragraphs that could contain more details.

SEARCHING:

Old:
The state is exited when the PROBE_COUNT reaches MAX_PROBES, a received PTB message is validated, a probe of size MAX_PMTU is acknowledged, or a black hole is detected.

New:
The state is exited to enter SEARCH_COMPLETE when the PROBE_COUNT reaches MAX_PROBES, a validated PTB is received that reports a MTU equal to the last successfully probed size (PTB_SIZE = PLPMTU), or a probe of size MAX_PMTU is acknowledged (PLPMTU = MAX_PMTU).

The state is also left when a black hole is detected. This causes the PL sender to enter the BASE state.

SEARCH_COMPLETE:

Old:
DPLPMTUD remains in this state until either the PMTU_RAISE_TIMER expires, a received PTB message is validated, or a black hole is detected.

New:
DPLPMTUD remains in this state until either the PMTU_RAISE_TIMER expires or a black hole is detected. The former causes the PL sender to enter the SEARCHING state again in trying to raise PLPMTU. The latter causes the PL sender to enter the BASE state.

ERROR:

Old:
The state is exited when packet probes no longer detect the error or when the PL indicates that connectivity has been lost.

New:
The state is exited when packet probes no longer detect the error. The PL sender then enters the SEARCHING state.

@tuexen @adventureloop @gorryfair
Comments?

[adventureloop]

On Thu, Dec 19, 2019 at 03:50:45AM -0800, msvoelker wrote: Since #8 is closed now, I looked into our state diagram description again. I found a few paragraphs that could contain more details. **SEARCHING**: Old: The state is exited when the PROBE_COUNT reaches MAX_PROBES, a received PTB message is validated, a probe of size MAX_PMTU is acknowledged, or a black hole is detected. New: The state is exited to enter SEARCH_COMPLETE when the PROBE_COUNT reaches MAX_PROBES, a validated PTB is received that reports a MTU equal to the last successfully probed size (PTB_SIZE = PLPMTU), or a probe of size MAX_PMTU is acknowledged (PLPMTU = MAX_PMTU). The state is also left when a black hole is detected. This causes the PL sender to enter the BASE state. **SEARCH_COMPLETE**: Old: DPLPMTUD remains in this state until either the PMTU_RAISE_TIMER expires, a received PTB message is validated, or a black hole is detected. New: DPLPMTUD remains in this state until either the PMTU_RAISE_TIMER expires or a black hole is detected. The former causes the PL sender to enter the SEARCHING state again in trying to raise PLPMTU. The latter causes the PL sender to enter the BASE state. **ERROR**: Old: The state is exited when packet probes no longer detect the error or when the PL indicates that connectivity has been lost. New: The state is exited when packet probes no longer detect the error. The PL sender then enters the SEARCHING state. @tuexen @adventureloop @gorryfair Comments?

Seems good, I'll merge this with the other changes we have now.

…

-- You are receiving this because you were mentioned. Reply to this email directly or view it on GitHub: #9 (comment)

-- - [tj]

[msvoelker]

ERROR:

Old:
The state is exited when packet probes no longer detect the error or when the PL indicates that connectivity has been lost.

New:
The state is exited when packet probes no longer detect the error. The PL sender then enters the SEARCHING state.

The new text now is:
The state is exited when packet probes no longer detect the error or when the PL indicates that connectivity has been lost. The PL sender then enters the SEARCHING state.

We should remove "or when the PL indicates that connectivity has been lost", because we don't want to say that the PL sender enters SEARCHING, when the connectivity is lost.

[adventureloop]

merged, thanks