Errno 32 Broken pipe in Hydra worker log

[Smile172]

Hi,

I have set up the app, with a heavy forwarder, but sometimes data comes in, sometimes not.
I found some errors in the Hydra worker's log:
ERROR [ta_ontap_collection_worker://gamma:6012] [ProcessorPerfHandler] Problem collecting processor from server=netapp.domain : [Errno 32] Broken pipe
Traceback (most recent call last):
File "/opt/splunk/etc/apps/Splunk_TA_ontap/bin/ta_ontap/handlers.py", line 93, in runPerf
time=datetime.datetime.fromtimestamp(int(results[object_name]['timestamp'])))
File "/opt/splunk/etc/apps/SA-Hydra/bin/hydra/init.py", line 226, in sendData
self.out.flush()
IOError: [Errno 32] Broken pipe
ERROR [ta_ontap_collection_worker://gamma:6012] [MegaPerfHandler] failed sub job run on sub_handler= server=netapp.domain

Forwarder's OS version: Ubuntu 14.04 LTS

Thank you!

[Smile172 · Accepted solution]

halr9000, thank you for the answer!

After I disabled IPv6 on the heavy forwarder, the error message disappeared and the data flows seamlessly.

[Smile172]

Sometimes I have seen data in NetApp app sometimes not. At the beginning the heavy forwarder had 2 IP-s, 1 IPv4 and 1 IPv6 address. I disabled the IPv6 to force binding to IPv4 address and the connection restored 🙂

[halr9000]

Glad that worked out. What led you to that solution? I think some more detail in this answer would be helpful to others who may come after you with the same issue. TIA

[halr9000]

Can you try this on RHEL or CentOS? Those are the only supported platforms for a DCN according to the docs (http://docs.splunk.com/Documentation/NetApp/latest/DeployNetapp/Platformandhardwarerequirements#Splu...). If it's still broken there, then open a support case so that we can get your diag logs.