[PATCH] DLM: increase socket backlog to avoid hangs with 16 nodes
From: Edwin Török
Date: Thu Feb 02 2023 - 10:24:32 EST
On a 16 node virtual cluster with e1000 NICs joining the 12th node prints
SYN flood warnings for the DLM port:
Dec 21 01:46:41 localhost kernel: [ 2146.516664] TCP: request_sock_TCP: Possible SYN flooding on port 21064. Sending cookies. Check SNMP counters.
And then joining a DLM lockspace hangs:
```
Dec 21 01:49:00 localhost kernel: [ 2285.780913] INFO: task xapi-clusterd:17638 blocked for more than 120 seconds. │
Dec 21 01:49:00 localhost kernel: [ 2285.786476] Not tainted 4.4.0+10 #1 │
Dec 21 01:49:00 localhost kernel: [ 2285.789043] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. │
Dec 21 01:49:00 localhost kernel: [ 2285.794611] xapi-clusterd D ffff88001930bc58 0 17638 1 0x00000000 │
Dec 21 01:49:00 localhost kernel: [ 2285.794615] ffff88001930bc58 ffff880025593800 ffff880022433800 ffff88001930c000 │
Dec 21 01:49:00 localhost kernel: [ 2285.794617] ffff88000ef4a660 ffff88000ef4a658 ffff880022433800 ffff88000ef4a000 │
Dec 21 01:49:00 localhost kernel: [ 2285.794619] ffff88001930bc70 ffffffff8159f6b4 7fffffffffffffff ffff88001930bd10
Dec 21 01:49:00 localhost kernel: [ 2285.794644] [<ffffffff811570fe>] ? printk+0x4d/0x4f │
Dec 21 01:49:00 localhost kernel: [ 2285.794647] [<ffffffff810b1741>] ? __raw_callee_save___pv_queued_spin_unlock+0x11/0x20 │
Dec 21 01:49:00 localhost kernel: [ 2285.794649] [<ffffffff815a085d>] wait_for_completion+0x9d/0x110 │
Dec 21 01:49:00 localhost kernel: [ 2285.794653] [<ffffffff810979e0>] ? wake_up_q+0x80/0x80 │
Dec 21 01:49:00 localhost kernel: [ 2285.794661] [<ffffffffa03fa4b8>] dlm_new_lockspace+0x908/0xac0 [dlm] │
Dec 21 01:49:00 localhost kernel: [ 2285.794665] [<ffffffff810aaa60>] ? prepare_to_wait_event+0x100/0x100 │
Dec 21 01:49:00 localhost kernel: [ 2285.794670] [<ffffffffa0402e37>] device_write+0x497/0x6b0 [dlm] │
Dec 21 01:49:00 localhost kernel: [ 2285.794673] [<ffffffff811834f0>] ? handle_mm_fault+0x7f0/0x13b0 │
Dec 21 01:49:00 localhost kernel: [ 2285.794677] [<ffffffff811b4438>] __vfs_write+0x28/0xd0 │
Dec 21 01:49:00 localhost kernel: [ 2285.794679] [<ffffffff811b4b7f>] ? rw_verify_area+0x6f/0xd0 ┤
Dec 21 01:49:00 localhost kernel: [ 2285.794681] [<ffffffff811b4dc1>] vfs_write+0xb1/0x190 │
Dec 21 01:49:00 localhost kernel: [ 2285.794686] [<ffffffff8105ffc2>] ? __do_page_fault+0x302/0x420 │
Dec 21 01:49:00 localhost kernel: [ 2285.794688] [<ffffffff811b5986>] SyS_write+0x46/0xa0 │
Dec 21 01:49:00 localhost kernel: [ 2285.794690] [<ffffffff815a31ae>] entry_SYSCALL_64_fastpath+0x12/0x71
```
The previous limit of 5 seems like an arbitrary number, that doesn't match any
known DLM cluster size upper bound limit.
Signed-off-by: Edwin Török <edvin.torok@xxxxxxxxxx>
Cc: Christine Caulfield <ccaulfie@xxxxxxxxxx>
Cc: David Teigland <teigland@xxxxxxxxxx>
Cc: cluster-devel@xxxxxxxxxx
---
Notes from 2023:
This patch was initially developed on 21 Dec 2017, and in production use ever since.
I expected to drop out of our patchqueue at the next kernel upgrade, however it
hasn't, so I probably forgot to send it.
I haven't noticed this bug again with the patch applied, and the previous value
of '5' seems like an arbitrary limit not matching any supported upper bounds
on DLM cluster sizes, so this patch has (unintentionally) had a 5 year test
cycle.
Although the join hanging forever like that may still be a bug, if the SYN cookies
consistently trigger it lets try to avoid the bug by avoiding the SYN cookies.
---
fs/dlm/lowcomms.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/fs/dlm/lowcomms.c b/fs/dlm/lowcomms.c
index 4450721ec..105c9138b 100644
--- a/fs/dlm/lowcomms.c
+++ b/fs/dlm/lowcomms.c
@@ -1774,7 +1774,7 @@ static int dlm_listen_for_all(void)
sock->sk->sk_data_ready = lowcomms_listen_data_ready;
release_sock(sock->sk);
- result = sock->ops->listen(sock, 5);
+ result = sock->ops->listen(sock, 128);
if (result < 0) {
dlm_close_sock(&listen_con.sock);
return result;
--
2.34.1