#Change default port 443 to talk with ontap cluster

1 messages · Page 1 of 1 (latest)

raven tinsel
#

Hello,
I don't know if I am wrong, but it sounds familiar to me that there were a way to attack a ONTAP cluster from harvest using a different port that default (443). I will try to explain our use case.
We have a service to redirect the traffic between the ONTAP cluster and our harvest machine which forces us to use a different port than 443 to send the request to storage array and talk with it, for that reason we need to specify in our harvest.yml that defined port for every specific cluster but others should still using the common configuration. So there the question, can we change the default 443 port by other to talk with storage array for every poller?

I hope is clear, if not please let me know to explain better.
Thanks in advance.

slow lynx
#

pretty sure you can.
You have to edit the harvest.yml file
Change the port range under pollers section
you can also use a different harvest file and add the --config command to the start command to test the config before changing it in the produciton one.
Also would need to change the poller port on each cluster via the addr: $cluster-management-ip:xxxx

dont think you need to do anything with the grafana or prom config, since you aren't changing those ports, but I am not 100% on that

#

if you have a really old version, i think the config file was netapp-worker, but that's really old and i dont remember what settings in that file need changed

#

validate by using
http://<poller-ip>:XXXX/
if not mistaken

thick osprey
#

@raven tinsel in your harvest.yml file, you can specify the ONTAP port by appending it to your addr value. The following example connects to ONTAP on port 9943

Pollers:
    sar: 
        datacenter: dc-1
        addr: 10.193.48.154:9943
raven tinsel
#

Thanks for your quick response.
I have tried to configure as you show in the example but something isn't working well. I don't know if could be related with the ONTAP version, if there is something different when we want monitor a fsx than a physical storage array.

I am seeing some errors but they are related with hardware capacity like this (DIMM), which it makes sense for me ins this particular case:

time=2026-02-23T18:15:03.004+01:00 level=ERROR source=collector.go:436 msg="" Poller=xxxxxxx collector=Rest:DIMM error="failed to fetch data: error making request StatusCode: 404, Message: API not found, Code: 3, API: /api/private/cli/system/controller/fru?fields=fru_name%2Cnode%2Cserial_number%2Cstatus%2Csubsystem&ignore_unknown_fields=true&max_records=500&return_records=true&subsystem=Memory" task=data

There are other errors for using ZAPI and I think this cluster only use REST-API but in the collectors we have all of them enabled but since we still have clusters in older versions, firstly, the collectors for ZAPI are enabled.

time=2026-02-23T18:13:02.176+01:00 level=ERROR source=zapi.go:506 msg="Unable to get nodes" Poller=xxxxx collector=Zapi:Node error="failure invoking zapi: system-node-get-iter connection error => Post \"https://xxxxx:43011:443/servlets/netapp.servlets.admin.XMLrequest_filer\": dial tcp: lookup xxxx:43011: no such host"

thick osprey
#

Which version of Harvest are you using?

raven tinsel
thick osprey
raven tinsel
thick osprey
#

thanks!

raven tinsel
#

Hello Chris,
I have updated the harvest versión to latest: 26.02 but the error about DIMM, FRU still happening. Additional to those I have seen other errors related to Security (certificates) but it is because the harvest try to communicate with ZAPI for this cluster.

time=2026-02-24T12:09:52.944+01:00 level=ERROR source=collector.go:556 msg="export data" Poller=xxxxxx collector=Zapi:SecurityCert error="unable to emit object: security_certificate, uuid: Zapi, err=API request rejected: {\"error\":\"partial write: field type conflict: input field \\\"expiry_time\\\" on measurement \\\"security_certificate\\\" is type float, already exists as type string dropped=8\"}\n" exporter=influx

The DIMM, FRU collectors:

time=2026-02-24T12:15:50.980+01:00 level=ERROR source=collector.go:449 msg="" Poller=xxxxxxx collector=Rest:**DIMM** error="failed to fetch data: error making request StatusCode: 404, Message: API not found, Code: 3, API: /api/private/cli/system/controller/fru?fields=fru_name%2Cnode%2Cserial_number%2Cstatus%2Csubsystem&ignore_unknown_fields=true&max_records=500&return_records=true&subsystem=Memory" task=data

time=2026-02-24T12:15:51.818+01:00 level=ERROR source=collector.go:449 msg="" Poller=xxxxxx collector=Rest:**FRU** error="failed to fetch data: error making request StatusCode: 404, Message: API not found, Code: 3, API: /api/private/cli/system/controller/fru?fields=fru_name%2Cnode%2Cserial_number%2Cstatus%2Csubsystem&ignore_unknown_fields=true&max_records=500&return_records=true" task=data

thick osprey
#

Glad that the port fix worked for you! The 404's are expected for FSx because it does not include APIs for those endpoints. The influxdb error looks like a bug. We'll take a look and get back to you

raven tinsel
#

Thanks chris

thick osprey
#

Hi @raven tinsel the influxdb error is fixed in the latest nightly build if you want to give it a try

raven tinsel