1.9.2: Crash with 300% CPU and stuck agent-checks

Hello, We faced this evening a critical issue this issue where all agent-checks ware stuck (or retries very very slower than usual). In example I see "2h39m DOWN 6/15 ↑" for more than 2h one several backend servers. So all down server stayed down until I manually forced "UP" state. The problem se

Re: 1.9.2: Crash with 300% CPU and stuck agent-checks

January 29, 2019 4:24:57 AM CET Willy Tarreau wrote:Hello Louis, On Mon, Jan 28, 2019 at 10:43:37PM +0100, Louis Chanouha wrote: > Hello, > We faced this evening a critical issue this issue where all agent-checks ware > stuck (or retries very very slower than usual). > In example

Re: 1.9.2: Crash with 300% CPU and stuck agent-checks

8 cpust_ms_tot: 6731598 6509804 640930 cpust_ms_1s: 313 219 3 cpust_ms_15s: 3864 3003 142 avg_loop_us: 1 0 43 January 29, 2019 10:45:58 AM CET Willy Tarreau wrote:On Tue, Jan 29, 2019 at 10:41:52AM +0100, Louis Chanouha wrote: > I'm pretty sure this bug is specific to version 1.9. Last week

Re: 1.9.2: Crash with 300% CPU and stuck agent-checks

er_thread[2].task_list_size $4 = 0 (gdb) p task_per_thread[3].task_list_size $5 = 0 It will help ruling out certain areas which could set the negative value. Thanks, Willy -- Louis Chanouha | Infrastructures informatiques Service Numérique de l'Université de Toulouse Université Fé

Re: 1.9.2: Crash with 300% CPU and stuck agent-checks

].task_list_size cannot subscript something of type `' (gdb) down #0  0x7f1549f8e303 in epoll_wait () at ../sysdeps/unix/syscall-template.S:84 84      in ../sysdeps/unix/syscall-template.S (gdb) p task_per_thread[0].task_list_size cannot subscript something of type `' March 14, 2019 2

Setting retry-on causes segmentation fault with TCP backends

Louis --- -- Louis Chanouha | Infrastructures informatiques Service Numérique de l'Université de Toulouse Université Fédérale Toulouse Midi-Pyrénées Maison de la Recherche et de la Valorisation - MRV 118 route de Narbonne - 31062 Toulouse Cedex 09 Tél. : +33 5 61 10 80 45 /poste int.

[bug] http-reuse and TCP mode warning when using PROXY protocol

Hello, I set a global http-reuse safe. HAProxy displays a warning for http-reuse and send-proxy combinaison on TCP mode backends, but http-reuse is active only on HTTP mode backends. [WARNING] 115/135122 (26529) : config : proxy ' SMTPS_SUBMISSION' : connections to server 'f1' will have a PR

h2 & server PUSH

Hello, If I'm right (I may have missed some exchanges in mailing), h2 main improvement in 1.9  will be end2end working. So to have an h2 with Server Push, we will need to have h2 enabled backends. Is a server push initiated by HAProxy based on "Link" header scheduled to 1.9 (like nginx http2