Howdy,

I've noticed a bug occuring in the Event Editor on the RPi.

I had put in all the hub polls in a nice ordered list 1 minute apart.

After a while I notice;

1) The order of the events changes
2) Some events are deleted and some events are duplicated

This is what I see now;

Poll SpookNet HUB 00:05
Poll FidoNet HUB 00:02
Poll RetroNet HUB 00:03
Poll SpookNet HUB 00:04
Poll DoreNet HUB 00:07
Poll SportNet HUB 00:10
Poll PiNet HUB 00:08
Poll JustaXNet HUB 00:09
Poll SysopTechNet HUB 00:13
Poll fsxNET Hub 00:12
Poll MicroNet HUB 00:15
Poll GameNet HUB 00:18
Poll AdventureNet HUB 00:17
Poll VKRadio HUB 00:18
Poll GameNet HUB 00:18
Poll CommodoreNet HUB 00:19

Like I said, each poll was set up to poll on a seperate minute so they
wouldn't clash, and they were set up in order 00:01, 00:02, 00:03 etc...

--- Mystic BBS v1.12 A39 2018/04/21 (Raspberry Pi/32)
* Origin: Fusion BBS ~ Newcastle, Australia (21:4/132)

On 12/13/18, Michael Borthwick pondered and said...

I've noticed a bug occuring in the Event Editor on the RPi.

I had put in all the hub polls in a nice ordered list 1 minute apart.

Can you show the settings of one of the event entries please?
Is this an interval or hourly style of poll?

--- Mystic BBS v1.12 A40 2018/12/11 (Windows/32)
* Origin: Agency BBS | Dunedin, New Zealand | agency.bbs.nz (21:1/101)

Can you show the settings of one of the event entries please?
Is this an interval or hourly style of poll?

It's an hourly poll like set up in the videos;

Active: Yes
Description: Poll SpookNet HUB
Exec Type: Hourly
Exec Hour: 0
Exec Min: 5
Shell: ./fidopoll 700:100/0|./mutil ImportEchoMail.ini
Semaphore: <blank>
Kill After: <blank>
Warning: 0
Sun: Yes
Mon: Yes
Tue: Yes
Wed: Yes
Thu: Yes
Fri: Yes
Sat: Yes

They are all identical except for the description, exec min, and the shell fidopoll address.

--- Mystic BBS v1.12 A39 2018/04/21 (Raspberry Pi/32)
* Origin: Fusion BBS ~ Newcastle, Australia (21:4/132)

On 12/13/18, Michael Borthwick pondered and said...

Poll AdventureNet HUB 00:17
Poll VKRadio HUB 00:18
Poll GameNet HUB 00:18
Poll CommodoreNet HUB 00:19

Like I said, each poll was set up to poll on a seperate minute so they wouldn't clash, and they were set up in order 00:01, 00:02, 00:03 etc...

Thanks for the posted info. I wonder if what is happening is that because you have each of these stacked so closely next to each other, if one fails to complete running prior to the due start time of the next (only 60 secs apart) then the future run times of all are knocked out of kilter.

Suggest you add extra time between each, say 5 mins apart and see if that solves it.

--- Mystic BBS v1.12 A40 2018/12/11 (Windows/32)
* Origin: Agency BBS | Dunedin, New Zealand | agency.bbs.nz (21:1/101)

I was actually thinking of running these events in cron instead and only keeping the semaphore events in Mystic.

--- Mystic BBS v1.12 A39 2018/04/21 (Raspberry Pi/32)
* Origin: Fusion BBS ~ Newcastle, Australia (21:4/132)

On 12/13/18, Michael Borthwick pondered and said...

I was actually thinking of running these events in cron instead and only keeping the semaphore events in Mystic.

Yep whatever works. But calling them via cron may not solve an issue of
things finishing on time and in the case of fidopoll you can only run one instance at a time. So the last event called does need to finish first before you run the next.

You could also just run

'./fidopoll forced ' to poll them all in one sequential effort :)

--- Mystic BBS v1.12 A40 2018/12/11 (Windows/32)
* Origin: Agency BBS | Dunedin, New Zealand | agency.bbs.nz (21:1/101)

Yep whatever works. But calling them via cron may not solve an issue of things finishing on time and in the case of fidopoll you can only run one instance at a time. So the last event called does need to finish first before you run the next.

I was thinking of running via a script to check if fidopoll was running, if
not run fidopoll, if it is then just abort.

--- Mystic BBS v1.12 A39 2018/04/21 (Raspberry Pi/32)
* Origin: Fusion BBS ~ Newcastle, Australia (21:4/132)

Something along the lines of;

#!/bin/bash
MDIR=/extdrive/mystic
FBSY=/extdrive/mystic/semaphore/fidopoll.bsy
if [-f "$MBSY" ];
then
exit 1
else
cd $MDIR
./fidopoll 21:4/100
./mutil ImportEchoMail.ini
fi

--- Mystic BBS v1.12 A39 2018/04/21 (Raspberry Pi/32)
* Origin: Fusion BBS ~ Newcastle, Australia (21:4/132)

Easy stuff so now cron handles my backup and nightly maintenance in one at 00:00 and then from 01:00 to 23:00 runs the fidopoll scripts.. the only thing Mystic does now is handle the semaphores.. so fingers crossed :)

--- Mystic BBS v1.12 A39 2018/04/21 (Raspberry Pi/32)
* Origin: Fusion BBS ~ Newcastle, Australia (21:4/132)

On 12-13-18 15:24, Michael Borthwick wrote to Avon <=-

I was actually thinking of running these events in cron instead and
only keeping the semaphore events in Mystic.

You will still have semaphore issues, because these are managed by Mystic/fidopoll. Spacing your poll events (so there's time to toss mail as well) is the way to go.


... Honk if you love peace and quiet.
=== MultiMail/Win v0.51
--- SBBSecho 3.03-Linux
* Origin: Freeway BBS Bendigo,Australia freeway.apana.org.au (21:1/109)

On 12-13-18 18:28, Michael Borthwick wrote to Avon <=-

Yep whatever works. But calling them via cron may not solve an issue of things finishing on time and in the case of fidopoll you can only run one instance at a time. So the last event called does need to finish first before you run the next.

I was thinking of running via a script to check if fidopoll was
running, if not run fidopoll, if it is then just abort.

It's not just fidopoll, it could be mutil tossing the previous batch of mail that gums up the works too. :)


... Useless Invention: Ejector seats for helicopters.
=== MultiMail/Win v0.51
--- SBBSecho 3.03-Linux
* Origin: Freeway BBS Bendigo,Australia freeway.apana.org.au (21:1/109)

It's not just fidopoll, it could be mutil tossing the previous batch of mail that gums up the works too. :)

I didn't think of that, I'll just pop in another check to see if mutil is running as well.

--- Mystic BBS v1.12 A39 2018/04/21 (Raspberry Pi/32)
* Origin: Fusion BBS ~ Newcastle, Australia (21:4/132)

On 12-14-18 10:00, Michael Borthwick wrote to Vk3jed <=-

It's not just fidopoll, it could be mutil tossing the previous batch of mail that gums up the works too. :)

I didn't think of that, I'll just pop in another check to see if mutil
is running as well.

Hmm, if they get backed up, I see potential for race conditions if there's 2 or more instances waiting. If the scripts simply abort, you should be OK... I think. :D But then, some may never get to run, if the timing is wrong.


... MISSING: Tagline, 70 characters long, last seen in New Mexico.
=== MultiMail/Win v0.51
--- SBBSecho 3.03-Linux
* Origin: Freeway BBS Bendigo,Australia freeway.apana.org.au (21:1/109)

It's not just fidopoll, it could be mutil tossing the previous batch of mail that gums up the works too. :)

Ok so this should do it;

#!/bin/sh
MDIR=/extdrive/mystic
FBSY=/extdrive/mystic/semaphore/fidopoll.bsy MBSY=/extdrive/mystic/semaphore/mutil.bsy

if [-f "$FBSY" ];
then
exit 1
else
if [-f "$MBSY" ];
then
exit 1
else
cd $MDIR
./fidopoll 33:2100/1
./mutil ImportEchoMail
fi
fi

--- Mystic BBS v1.12 A39 2018/04/21 (Raspberry Pi/32)
* Origin: Fusion BBS ~ Newcastle, Australia (21:4/132)

Hmm, if they get backed up, I see potential for race conditions if
there's 2 or more instances waiting. If the scripts simply abort, you should be OK... I think. :D But then, some may never get to run, if the timing is wrong.

The scripts do an abort, they don't wait X seconds and try again.

I think 1 minute is plenty of time to give a poll and import given the
message loads of today. Unless someone is trying to send a massive (many many MB). If that's the case each script will just abort and rerun next hour. It's not hard to space them out to 2 or 3 minutes if necessary.

--- Mystic BBS v1.12 A39 2018/04/21 (Raspberry Pi/32)
* Origin: Fusion BBS ~ Newcastle, Australia (21:4/132)

On 12/14/18, Vk3jed said the following...

... MISSING: Tagline, 70 characters long, last seen in New Mexico.

I didn't take it... Pequito, do you have this missing tagline? ;)


---

Black Panther(RCS)
a.k.a. Dan Richter
Sysop - Castle Rock BBS
telnet://bbs.castlerockbbs.com
http://www.castlerockbbs.com
The sparrows are flying again....

--- Mystic BBS v1.12 A39 2018/04/21 (Linux/64)
* Origin: Castle Rock BBS - bbs.castlerockbbs.com (21:1/186)

On 12-14-18 10:07, Michael Borthwick wrote to Vk3jed <=-

It's not just fidopoll, it could be mutil tossing the previous batch of mail that gums up the works too. :)

Ok so this should do it;

Should do the trick. :)


... For a good time call 86753099 (Jenny)...
=== MultiMail/Win v0.51
--- SBBSecho 3.03-Linux
* Origin: Freeway BBS Bendigo,Australia freeway.apana.org.au (21:1/109)

On 12-14-18 13:06, Michael Borthwick wrote to Vk3jed <=-

The scripts do an abort, they don't wait X seconds and try again.

I think 1 minute is plenty of time to give a poll and import given the message loads of today. Unless someone is trying to send a massive
(many many MB). If that's the case each script will just abort and
rerun next hour. It's not hard to space them out to 2 or 3 minutes if necessary.

That makes a few assumptions:

1. That there's no undue delays in polling (not gusranteed when running IPv4 and IPv6)

2. That tossing happens immediately after polling (I'm not sure how often Mystic checks).

3. Any other unexpected delays.

I still prefer to leave a few minutes between polling events. :)


... Warranty: If it breaks, both halves are yours.
=== MultiMail/Win v0.51
--- SBBSecho 3.03-Linux
* Origin: Freeway BBS Bendigo,Australia freeway.apana.org.au (21:1/109)

On 12-13-18 19:22, Black Panther wrote to Vk3jed <=-

On 12/14/18, Vk3jed said the following...

... MISSING: Tagline, 70 characters long, last seen in New Mexico.

I didn't take it... Pequito, do you have this missing tagline? ;)

Hmm, thieves everywhere. :D


... I give up, what is the meaning of life?
=== MultiMail/Win v0.51
--- SBBSecho 3.03-Linux
* Origin: Freeway BBS Bendigo,Australia freeway.apana.org.au (21:1/109)

1. That there's no undue delays in polling (not gusranteed when running IPv4 and IPv6)

Fair enough, the internet isn't perfect.

2. That tossing happens immediately after polling (I'm not sure how
often Mystic checks).

I call the tosser manually after the fidopoll exits.

Mystic could call the tosser itself as well but the polling scripts will
check to see if that's running as well.

I'll space them another minute out, I don't know how many networks I'll end
up connecting to but every 2 minutes would limit me to 30 polls per hour. I guess when/if I reach 30 I would have to start running odd/even hours.

--- Mystic BBS v1.12 A39 2018/04/21 (Raspberry Pi/32)
* Origin: Fusion BBS ~ Newcastle, Australia (21:4/132)

On 12-14-18 17:10, Michael Borthwick wrote to Vk3jed <=-

1. That there's no undue delays in polling (not gusranteed when running IPv4 and IPv6)

Fair enough, the internet isn't perfect.

Indeed. :)

2. That tossing happens immediately after polling (I'm not sure how
often Mystic checks).

I call the tosser manually after the fidopoll exits.

That saves another uncertainty. Good. :)

I'll space them another minute out, I don't know how many networks I'll end up connecting to but every 2 minutes would limit me to 30 polls per hour. I guess when/if I reach 30 I would have to start running odd/even hours.

I doubt you'll reach 30. :)


... He's dead, Jim; kick him yourself if you don't believe me.
=== MultiMail/Win v0.51
--- SBBSecho 3.03-Linux
* Origin: Freeway BBS Bendigo,Australia freeway.apana.org.au (21:1/109)

It's not just fidopoll, it could be mutil tossing the previous batch mail that gums up the works too. :)

I found the semaphore events works well, but a good way to configure them is like:

(echomail.out|netmail.out|newsmail.out)!fidopoll.bsy

Same thing for inbound, but use mutil.bsy at the end.

That way, when semaphores are found but fidopoll or mutil is running, it will not delete the semaphores and run then fidopoll or mutil when it's done.

Also, I do not run the import after fidopoll, if there is new messages received, the semaphore will get created and the semaphore event will pick
them up, no need to forcefully run mutil.

Cheers!

|15 ß Þ |15StackFault |08<|03.|11.|15P|11h|03EN|11o|15M|11.|03.|08>
|11 Ý ß |11The Bottomless Abyss BBS
|03 ß Ýß |03ssh|08.|072222 |08/ |03telnet|08.|072023 |08/ |03https
|08 ÜþÞ |08bbs|07.|08bottomlessabyss|07.|08net

--- Mystic BBS v1.12 A39 2018/04/21 (Linux/64)
* Origin: The Bottomless Abyss BBS þ bbs.bottomlessabyss.net (21:1/172)

On 12/14/18, Michael Borthwick said the following...

I'll space them another minute out, I don't know how many networks I'll end up connecting to but every 2 minutes would limit me to 30 polls per hour. I guess when/if I reach 30 I would have to start running odd/even hours.

What I did here, is set up a script to run my fidopolls. Every hour it'll run it and poll each of them.

#!/bin/sh
./fidopoll 21:1/100
./fidopoll 3:770/1
etc...

Then I have it set to toss all the mail at once when it's done.


---

Black Panther(RCS)
a.k.a. Dan Richter
Sysop - Castle Rock BBS
telnet://bbs.castlerockbbs.com
http://www.castlerockbbs.com
The sparrows are flying again....

--- Mystic BBS v1.12 A39 2018/04/21 (Linux/64)
* Origin: Castle Rock BBS - bbs.castlerockbbs.com (21:1/186)