I noticed an issue on my OpenSIPS B2BUA server where Heartbeat gave me the following errors
<br />
<br />Jun 21 11:50:46 proxy01 logd: [1086]: WARN: G_CH_check_int: working on IPC channel took 160 ms (> 100 ms)
<br />Jun 21 11:50:48 proxy01 logd: [1084]: WARN: G_CH_check_int: working on IPC channel took 770 ms (> 100 ms)
<br />Jun 21 11:50:48 proxy01 logd: [1084]: WARN: G_CH_check_int: working on IPC channel took 370 ms (> 100 ms)
<br />Jun 21 11:50:48 proxy01 logd: [1086]: WARN: G_CH_prepare_int: working on IPC channel took 390 ms (> 100 ms)
<br />Jun 21 11:50:48 proxy01 logd: [1084]: WARN: G_CH_check_int: working on IPC channel took 140 ms (> 100 ms)
<br />Jun 21 11:50:48 proxy01 logd: [1084]: WARN: G_CH_prepare_int: working on IPC channel took 130 ms (> 100 ms)
<br />Jun 21 11:50:48 proxy01 crmd: [21371]: WARN: G_CH_check_int: working on IPC channel took 290 ms (> 100 ms)
<br />Jun 21 11:50:48 proxy01 ccm: [21366]: WARN: G_CH_check_int: working on IPC channel took 540 ms (> 100 ms)
<br />Jun 21 11:50:48 proxy01 lrmd: [21368]: WARN: G_CH_check_int: working on IPC channel took 650 ms (> 100 ms)
<br />Jun 21 11:50:48 proxy01 cib: [21367]: WARN: G_CH_check_int: working on Heartbeat API channel took 1630 ms (> 100 ms)
<br />Jun 21 11:50:48 proxy01 crmd: [21371]: WARN: G_CH_check_int: working on IPC channel took 350 ms (> 100 ms)
<br />Jun 21 11:50:48 proxy01 crmd: [21371]: WARN: G_CH_check_int: working on IPC channel took 220 ms (> 100 ms)
<br />Jun 21 11:50:49 proxy01 cib: [21367]: WARN: G_CH_check_int: working on IPC channel took 220 ms (> 100 ms)
<br />Jun 21 11:50:50 proxy01 cib: [21367]: WARN: G_CH_check_int: working on IPC channel took 170 ms (> 100 ms)
<br />
<br />the errors eventually went away without me having to restart anything but because of Heartbeat it was causing my SIP Trunk provider to send multiple INVITES when a ITSP Caller was calling one of my users because the OpenSIPS B2BUA server took so long to reply to the INVITEs. Because a ton of INVITES were sent the OpenSIPS B2BUA sends a ton of INVITES to the OpenSIPS Proxy and then the proxy sends a ton of calls to the UA client. So the client gets bombed with a ton of calls.
<br />
<br />Is there any way from an OpenSIPS perspective to protect myself from from this? I currently can set each user to only have a certain amount of call channels, but this would just cause the extra channels to get loaded up. I was hoping there was a way to notice multiple INVITES with the same CALL-ID coming in and only relay the single INVITE on to the Proxy.
<br />
<br />Anyone else run into this?