Comment 21 for bug 802850

Revision history for this message
Mike Willbanks (mike-digitalstruct) wrote :

After a restart of all of the workers - the process through strace for gearmand is far, far, far different:

[pid 14267] 22:16:37.171837 [ 3692a0d605] futex(0x1838fc28, FUTEX_WAKE_PRIVATE, 1 <unfinished ...>
[pid 14269] 22:16:37.171975 [ 3691ed4f41] <... recvfrom resumed> 0x183b3de8, 8192, 64, 0, 0) = -1 EAGAIN (Resource temporarily unavailable) <0.000134>
[pid 14267] 22:16:37.172017 [ 3692a0d605] <... futex resumed> ) = 0 <0.000153>
[pid 14269] 22:16:37.172118 [ 3691ed48a8] epoll_wait(21, <unfinished ...>
[pid 14267] 22:16:37.172158 [ 3691ec680b] write(25, "\4", 1 <unfinished ...>
[pid 14269] 22:16:37.172267 [ 3691ed48a8] <... epoll_wait resumed> {{EPOLLIN, {u32=24, u64=24}}}, 32, 4294967295) = 1 <0.000111>
[pid 14267] 22:16:37.172307 [ 3691ec680b] <... write resumed> ) = 1 <0.000119>
[pid 14269] 22:16:37.172402 [ 36932044eb] clock_gettime(CLOCK_MONOTONIC, <unfinished ...>
[pid 14267] 22:16:37.172443 [ 3692a0aee9] futex(0x1838fc54, FUTEX_WAIT_PRIVATE, 3690998537, NULL <unfinished ...>
[pid 14269] 22:16:37.172545 [ 36932044eb] <... clock_gettime resumed> {7654352, 151281722}) = 0 <0.000105>
[pid 14269] 22:16:37.172638 [ 3691ec678b] read(24, "\4", 256) = 1 <0.000058>
[pid 14269] 22:16:37.172798 [ 3691ed5151] sendto(33, "\0RES\0\0\0\n\0\0\0\0", 12, MSG_DONTWAIT|MSG_NOSIGNAL, NULL, 0) = 12 <0.000074>
[pid 14269] 22:16:37.172986 [ 3691ec678b] read(24, 0x435fff40, 256) = -1 EAGAIN (Resource temporarily unavailable) <0.000057>
[pid 14269] 22:16:37.173144 [ 3691ed48a8] epoll_wait(21, {{EPOLLIN, {u32=33, u64=33}}}, 32, 4294967295) = 1 <0.000086>
[pid 14269] 22:16:37.173343 [ 36932044eb] clock_gettime(CLOCK_MONOTONIC, {7654352, 152178722}) = 0 <0.000059>
[pid 14269] 22:16:37.173507 [ 3691ed4f41] recvfrom(33, "\0REQ\0\0\0\4\0\0\0\0", 8192, MSG_DONTWAIT, NULL, NULL) = 12 <0.000056>
[pid 14269] 22:16:37.173674 [ 3692a0b316] futex(0x1838fc54, FUTEX_WAKE_OP_PRIVATE, 1, 1, 0x1838fc50, {FUTEX_OP_SET, 0, FUTEX_OP_CMP_GT, 1}) = 1 <0.000067>
[pid 14267] 22:16:37.173793 [ 3692a0aee9] <... futex resumed> ) = 0 <0.001323>
[pid 14269] 22:16:37.173883 [ 3691ed4f41] recvfrom(33, <unfinished ...>
[pid 14267] 22:16:37.173924 [ 3692a0d605] futex(0x1838fc28, FUTEX_WAKE_PRIVATE, 1 <unfinished ...>
[pid 14269] 22:16:37.174023 [ 3691ed4f41] <... recvfrom resumed> 0x183b3de8, 8192, 64, 0, 0) = -1 EAGAIN (Resource temporarily unavailable) <0.000102>
[pid 14267] 22:16:37.174061 [ 3692a0d605] <... futex resumed> ) = 0 <0.000109>
[pid 14269] 22:16:37.174156 [ 3691ed48a8] epoll_wait(21, <unfinished ...>
[pid 14267] 22:16:37.174198 [ 3692a0aee9] futex(0x1838fc54, FUTEX_WAIT_PRIVATE, 3690998539, NULL <unfinished ...>

This really looks like there is a much larger issue based on how this would normally react.