lapin: Heartbeat blocks all consumers from continuing in 0.12

MeterSnapshot { count: 414665, rates: [390.8100698087718, 730.5750441311204, 963.6300314729459], mean: 813.0564690986009 }
name: message_count labels: None
MeterSnapshot { count: 420872, rates: [394.6336229561739, 715.0688686739221, 954.5356931002789], mean: 801.6491752048053 }
name: message_count labels: None
MeterSnapshot { count: 425793, rates: [379.84412870857903, 696.1875593463596, 944.1788324173962], mean: 788.4941713438777 }
name: message_count labels: None
MeterSnapshot { count: 431172, rates: [374.9080306879829, 679.7197843340775, 934.5018666347837], mean: 776.8754486373127 }
name: message_count labels: None
MeterSnapshot { count: 436343, rates: [368.5097618284864, 663.3929785381047, 924.7543420213268], mean: 765.5033310586701 }
DEBUG 2018-06-16T19:19:38Z: lapin_futures::client: poll heartbeat
DEBUG 2018-06-16T19:19:38Z: lapin_futures::client: Sending heartbeat
name: message_count labels: None
MeterSnapshot { count: 439761, rates: [334.91339771824613, 642.0454100051794, 913.2244909211752], mean: 751.7178407188896 }
name: message_count labels: None
MeterSnapshot { count: 439761, rates: [260.830816404075, 610.7324858625518, 898.1302177517547], mean: 732.9250308062573 }
name: message_count labels: None
MeterSnapshot { count: 439761, rates: [203.13524406464748, 580.9467110509256, 883.285430973444], mean: 715.0489367894609 }
name: message_count labels: None
MeterSnapshot { count: 439761, rates: [158.20188714694842, 552.6136056185545, 868.6860069389069], mean: 698.0240798297584 }
name: message_count labels: None
MeterSnapshot { count: 439761, rates: [123.20775359341746, 525.6623220438021, 854.3278901587024], mean: 681.7910729352898 }
name: message_count labels: None
MeterSnapshot { count: 439761, rates: [95.95429497902224, 500.0254680794349, 840.2070921747342], mean: 666.2959256376611 }

These are my logs, showing 15 second windows of the rates. You can see that the moment the heartbeat sends, the count stops increasing and no messages are received. I confirmed this by looking in the RabbitMQ UI, and seeing a delivery rate of 0/s.

About this issue

Original URL
State: closed
Created 6 years ago
Comments: 18 (8 by maintainers)

Most upvoted comments

The only case where I am able to reproduce is if there are 0 message in the queue when the heartbeat is sent, then some messages get queued. In this case the new messages are only consumed when the next heartbeat comes.

Keruspe on Jun 25, 2018

I might have an idea. Let me polish what I have locally (that is finally working), that should fix it On Sun 17 Jun 2018 at 20:11, Isaac Whitfield notifications@github.com wrote:

It appears to be exactly that; a heartbeat will block the consumers - but when the next heartbeat sends, it unblocks and consumers start again. Very strange :p

— You are receiving this because you were mentioned.

Reply to this email directly, view it on GitHub https://github.com/sozu-proxy/lapin/issues/125#issuecomment-397896243, or mute the thread https://github.com/notifications/unsubscribe-auth/AANm3qAPLlB_OCPVO2vgg3iyRjI0gZk1ks5t9pvKgaJpZM4Uqlnh .

Keruspe on Jun 17, 2018