[SERVER-29920] restore the "server is now listening on port xxxx" log message Created: 29/Jun/17  Updated: 30/Oct/23  Resolved: 29/Jul/17

Status: Closed
Project: Core Server
Component/s: Networking
Affects Version/s: None
Fix Version/s: 3.5.11

Type: Bug Priority: Critical - P2
Reporter: Eric Milkie Assignee: Andrew Morrow (Inactive)
Resolution: Fixed Votes: 1
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Backwards Compatibility: Fully Compatible
Operating System: ALL
Sprint: Platforms 2017-07-10, Platforms 2017-07-31
Participants:

 Description   

It used to say something like "NETWORK [thread1] waiting for connections on port 27017" when startup was complete.



 Comments   
Comment by Githook User [ 29/Jul/17 ]

Author:

{'email': 'acm@mongodb.com', 'username': 'acmorrow', 'name': 'Andrew Morrow'}

Message: SERVER-29920 Restore missing message once listening

Also, fix some broken parameter handling for the legacy transport
Branch: master
https://github.com/mongodb/mongo/commit/a402b45dfb18951a2c0de17d52e2f8dabc01b5d2

Comment by Andrew Morrow (Inactive) [ 25/Jul/17 ]

milkie - I don't really see how it could be related. The dbclient networking still flows through the MessagingPort subsystem, which we haven't changed to use ASIO. jbreams please let me know if you disagree and can see a way your recent changes would have led to the observed changes in dbclient.

Comment by Eric Milkie [ 19/Jul/17 ]

Sorry to add on here, but I noticed the following and am wondering if they are all related to the ASIO changes..

dbclient.cpp can now report the following messages: (typically via the shell)

[js_test:change_notification] 2017-07-19T08:45:58.250-0400 2017-07-19T08:45:58.250-0400 I NETWORK  [thread1] reconnect 127.0.0.1:20010 (127.0.0.1) failed failed

and

[js_test:change_notification] 2017-07-19T08:45:58.378-0400 2017-07-19T08:45:58.378-0400 I NETWORK  [thread1] trying reconnect to 127.0.0.1:20010 (127.0.0.1) failed

Neither of which are ideal. The weirdness is coming from the fact that the errmsg string is being set to literally "failed" with no other context.

Comment by Andrew Morrow (Inactive) [ 29/Jun/17 ]

We will definitely get this addressed.

Comment by Andy Schwerin [ 29/Jun/17 ]

+1 for a log message when the server starts listening.

Comment by Andrew Morrow (Inactive) [ 29/Jun/17 ]

If you start up the server with the legacy transport layer, you will still get that message, but the new on-by-default ASIO based transport layer doesn't issue it.

Comment by James Wahlin [ 29/Jun/17 ]

I believe this is the code that generated this message: https://github.com/mongodb/mongo/blob/943ede3296b190c97fe6ae53f7b718de22cd074d/src/mongo/util/net/listen.cpp#L590

Generated at Thu Feb 08 04:22:09 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.