This was very easy to add because the DBselect() function already supportes an offset parameter.

See also ZBXNEXT-1503

support offset parameter history.get() for zabbix-3.0.2

Useful offset parameter.
Original attached patch for only zabix-2.2, and not fail patch for zabbix-3.0.2.

I rewrote the patch for zabbix-3.0.2. see attached file.
Please import this patch and support offset parameter.

Adding offset could lead to changing "Search/Filter elements limit" GUI configuration option to "Maximum rows per page".

Having pagenation support will help for other system integrations. When the historyget or items.get has more results its throwing 500 internal error. Pagination helps in api calls usecases

+1 It would also be useful in PowerShell to paging/piping the results in small badges. For example, we use PowerShell to synchronize the inventory data of all hosts with our CMDB via the API. But when you have 10,000 hosts, this can quickly lead to problems. Something like in SQL with offsets and limits would be great.

Are there any news regarding this feature? 
we are having troubles to sync zabbix hosts with our cmdb. Paging would be useful.

It's possible that the cause of this feature missing was support of things like "limit, offset" back in the days when we supported more database backends (like IBM DB2) which either don't have it or didn't have it back then. We now do not support IBM DB2 so the only question is Oracle.

According to this support for limiting and specifying offset was added to Oracle 12c. This should be fine since we require at least version 19c at this point.

I'll try to get more information.

simkimdm While this is still not implemented, would it work for you if you'd fetch the list of all hostids first and then go on handling them by smaller batches? Have you tried it?

@dimir works better than fetching all at once. thank you.
here is a small python snippet how i implemented it.

from zabbix_utils import ZabbixAPI
// ... login stuff
def get_zabbix_hosts():
    zabbix_hosts_ids = zabbix.host.get(
                                    output="hostid",
                                    evaltype=0,
                                    tags=[
                                       {
                                           "tag": "cmdb_id",
                                           "operator": 4
                                       }
                                    ]
    )
    chunk_size = 1000
    chunks = [chunk['hostid'] for chunk in zabbix_hosts_ids]
    zabbix_hosts = []
    for i in range(0, len(chunks), chunk_size):
        chunk = chunks[i:i + chunk_size]
        zabbix_hosts += zabbix.host.get(hostids=chunk,
                                        selectMacros="extend",
                                        selectTags="extend",
                                        selectHostGroups="extend",
                                        selectParentTemplates="extend",
                                        selectInterfaces="extend")
    return zabbix_hosts 

Nice. And nice to see you use our zabbix_utils and it works for you.