Interrupted system call

[azurit]

hi,

i'm monitoring 3 hosts (localhost + 2 remote hosts) and getting lots of 'server unreachable' false positives (also on localhost). i found this in server log:

Timeout while answering request
Get value from agent failed. Error: Cannot connect to [example.com:10050] [Interrupted system call]
Host [example]: first network error, wait for 15 seconds
Parameter [agent.ping] will be checked after 120 seconds on host [example]


any ideas what's wrong ? both server and agents are 1.4.3

[Niels]

It certainly looks like a network problem... Can the machines see each other? Try using Active Agent for your items.

[azurit]

It certainly looks like a network problem... Can the machines see each other? Try using Active Agent for your items.

it's definitely not network problem, everything other is working very good. i'm getting these false positives about 2 times per week (for random host and also _localhost_).

[scandic]

We're having this strange behavior, too. Every morning we're getting wrong alerts for 20 to 40 servers but those servers are up and running. Still no luck to find the bug here...

[starstream]

We see these problems to.

3618:20080319:101352 Get value from agent failed. Error: Cannot connect to [xxx.xxx.xxx.xxx:10050] [Interrupted system call]

We have no other problems with the network.

[Alexei]

I think that this can be caused by two reasons:

1. Insufficient number of trappers

Number of trappers, StartTrappers, defines how many requests ZABBIX can process simultaneously. It is safe to have hundreds of trappers running for large ZABBIX setups. Make sure that database engine accepts so many connections.

2. Bug in 1.4.4

I am pretty sure that the issue is already fixed in pre1.4.5 code two months ago.

[starstream]

Thanks for the fast reply as usual Alexei.

I will try adjusting the number of trappers.

I just need to optimize out database first, i just noticed it got over 40GB.

Number of hosts (monitored/not monitored/templates/deleted) 60(39/1/20/0)
Number of items (monitored/disabled/not supported)[trapper] 3688(2588/464/636)[0]
Number of triggers (enabled/disabled)[true/unknown/false] 1041(953/88)[0/6/947]

[yas]

Get value from agent failed

Get value from agent failed. Error: Cannot connect to [xxx.xxx.xxx.xxx:10050] [Interrupted system call]
IHAVE SAME PROBLEM ON MY SERVER