On 11/9/2014 3:11 PM, deoren wrote:
On 11/6/2014 7:36 AM, deoren wrote:
According to Pidgin, these are the modules I have active:
adhoc
admin_adhoc
announce
auth_dovecot
auto_accept_subscriptions
bosh
c2s
carbons
compression
disco
groups
http
iq
log_auth
mam
mam_archive
message
motd
offline
pep
ping
presence
private
proxy65
proxy65_whitelist
roster
s2s
saslauth
smacks
storage_internal
storage_sql2
time
tls
uptime
vcard
version
and if I'm running the right command (I'm new to Mercurial), these are
the changed files:
# hg status --rev 1513:1545
M mod_auth_pam/mod_auth_pam.lua
M mod_mam/mamprefsxml.lib.lua
M mod_mam_archive/mod_mam_archive.lua
M mod_mam_muc/mod_mam_muc.lua
M mod_muc_log/mod_muc_log.lua
M mod_profile/mod_profile.lua
M mod_smacks/mod_smacks.lua
A mod_roster_allinall/mod_roster_allinall.lua
A mod_sslv3_warn/mod_sslv3_warn.lua
R mod_websocket/mod_websocket.lua
R mod_websocket/websocket.html
When I first encountered this problem it was with r1535, so viewing just
those changed files:
# hg status --rev 1513:1535
M mod_mam/mamprefsxml.lib.lua
M mod_mam_archive/mod_mam_archive.lua
M mod_mam_muc/mod_mam_muc.lua
M mod_profile/mod_profile.lua
M mod_smacks/mod_smacks.lua
A mod_sslv3_warn/mod_sslv3_warn.lua
I looked back over the repo commit logs and compared against the modules
I'm running and only two applicable modules were modified between the
revision I was running (r1513) and where I ended up with the issues (r1535):
* mod_mam
* mod_smacks
I updated my local clone to r1514 and restarted Prosody. Overnight I had
none of the issues that plagued me previously (which would occur
somewhere within an 8 hour period after making the change).
It would appear that my issue is related to mod_smacks, either directly
or its interaction with another module.
I'm sure there is a better way to format this, but here are all of the
changes to mod_smacks that fall between the working and failing revision
points.
hg log -r1513:1535 mod_smacks | grep -E 'changeset|summary' | cut -f 2- -d '
' | sed -r 's/\s{2,}//g' | awk 'NR%2{printf $0 ;next;}1'
1517:8ac4438925cf mod_smacks: Examine the state of the stanza *before* it
was send
1518:9475fe14d58d mod_smacks: Correctly check if what is being sent is a
stanza
1520:2881d532f385 mod_smacks: Use filters for queuing outgoing stanzas
instead of wrapping session.send()
1522:d4a4ed31567e mod_smacks: Trick session.send() into not returning nil or
mod_message will act as if messages were not delivered
1526:120817435151 mod_smacks: Fix sending ack requests on s2s
1527:06ecc5b3ca46 mod_smacks: Send failure correctly on s2s
1528:5ccb3ee2cf72 mod_smacks: Remove variable not used after 2881d532f385
1529:16893646a458 mod_smacks: Move wrap_session() contents into separate
direction-specific routines
1530:fb7cd669f41b mod_smacks: Initiate outgoing smacks on s2s when sending
request and incoming when the remote says enabled
1531:7d86fc477993 mod_smacks: Delay initiation of outgoing smacks on s2s
until after queued stanzas are sent
My current plan is to jump to somewhere in the middle and see whether the
problem is present. I'd then start stepping backwards until it works again.
I ended up starting at r1528 and working backwards (restarting Prosody after
each run of hg update -rREVISION) each time the issues reappeared. I've made it
down to r1518 and so far it's been pretty stable. I had a client act pretty
peculiar (Yaxim) earlier today at r1518, but I think it may have been unrelated.
Is there anything further I can do to help troubleshoot the changes newer than
r1518?
--
You received this message because you are subscribed to the Google Groups
prosody-dev group.
To unsubscribe from this group and stop receiving emails from it, send an email
to prosody-dev+unsubscr...@googlegroups.com.
To post to this group, send email to prosody-dev@googlegroups.com.
Visit this group at http://groups.google.com/group/prosody-dev.
For more options, visit https://groups.google.com/d/optout.