AnsweredAssumed Answered

CTIOS errors - pending writes, exceeds limit.

Question asked by lukasz_wozniczka@sevenet.pl on Nov 14, 2014
Latest reply on Apr 15, 2015 by gcheria

Hi, Guys

 

In our environment, I discover multiple messages (related to one recorder) on CTIOS queue, they looks like cloned or multipled, example:

 

08:47:18:627 CTIOS1-ctios Trace: <-SND[07096][agent.5000.29327]:Monitored eAgentStateEvent(eTalking),Msk(0)

08:47:18:627 CTIOS1-ctios Trace: <-SND[07096][agent.5000.29327]:Monitored eAgentStateEvent(eTalking),Msk(0)

08:47:18:627 CTIOS1-ctios Trace: <-SND[07096][agent.5000.29327]:Monitored eAgentStateEvent(eTalking),Msk(0)

08:47:18:627 CTIOS1-ctios Trace: <-SND[07096][agent.5000.29327]:Monitored eAgentStateEvent(eTalking),Msk(0)

08:47:18:627 CTIOS1-ctios Trace: <-SND[07096][agent.5000.29327]:Monitored eAgentStateEvent(eTalking),Msk(0)

08:47:18:627 CTIOS1-ctios Trace: <-SND[07096][agent.5000.29327]:Monitored eAgentStateEvent(eTalking),Msk(0)

08:47:18:643 CTIOS1-ctios Trace: <-SND[07096][agent.5000.29327]:Monitored eAgentStateEvent(eTalking),Msk(0)

08:47:18:643 CTIOS1-ctios Trace: <-SND[07096][agent.5000.29327]:Monitored eAgentStateEvent(eTalking),Msk(0)

08:47:18:643 CTIOS1-ctios Trace: <-SND[07096][agent.5000.29327]:Monitored eAgentStateEvent(eTalking),Msk(0)

08:47:18:643 CTIOS1-ctios Trace: <-SND[07096][agent.5000.29327]:Monitored eAgentStateEvent(eTalking),Msk(0)

08:47:18:643 CTIOS1-ctios Trace: <-SND[07096][agent.5000.29327]:Monitored eAgentStateEvent(eTalking),Msk(0)

08:47:18:643 CTIOS1-ctios Trace: <-SND[07096][agent.5000.29327]:Monitored eAgentStateEvent(eTalking),Msk(0)

08:47:18:643 CTIOS1-ctios Trace: <-SND[07096][agent.5000.29327]:Monitored eAgentStateEvent(eTalking),Msk(0)

08:47:18:643 CTIOS1-ctios Trace: <-SND[07096][agent.5000.29327]:Monitored eAgentStateEvent(eTalking),Msk(0)

08:47:18:643 CTIOS1-ctios Trace: <-SND[07096][agent.5000.29327]:Monitored eAgentStateEvent(eTalking),Msk(0)

 

which leading to error:

 

08:47:18:549 CTIOS1-ctios Trace: IOCPConnectionMgrThread[6060],Client[07264, 05CE0600]:Pending Writes(8000) Exceeds Limit(8000)Client disconnected by CTIOS Server,Client(7264) at Address(X.X.X.X),Total Client pending write operations is (14464) for (121) Client(s) connected,LastWriteResponseTime(0 ms)

 

Assuming the queue exceeds more than 8000 pending writes, and then CTIOS treat the client as a slow client and closes the connection, I don't known what's the reason of this situation, and which cause may result in such behavior?

Do you have, any clue why queue exceeded pending writes (8000), and what can cause this (maybe wrong configuration of monitor mode's)?

 

I don't think it's normal behaviour, i mean whether there should be much less messages in that short range of time.

Outcomes