Hi, we added a column to a schema via PUT the sche...
# troubleshootinge
Elon
03/31/2021, 7:04 PMHi, we added a column to a schema via PUT the schema api (it's in the kafka topic) but still do not see data, i.e. the select list does not contain that column. Do we need to restart the servers or do some other api call?
Elon
03/31/2021, 7:04 PMOr do we have to wait for new segments?
x
Xiang Fu
03/31/2021, 7:28 PM
the segment created after the schema call should have new column
Xiang Fu
03/31/2021, 7:29 PM
Or you can reload the consuming segments or just restart pinot server
e
Elon
03/31/2021, 7:30 PMOh nice! How do you reload consuming segments?
Elon
03/31/2021, 7:41 PMAh, is it the "reload table if backwards compatible" checkbox in the swagger page?
x
Xiang Fu
03/31/2021, 7:42 PM
there is a swagger api to reload a segment. I haven’t used the UI button :p
👍 1
e
Elon
03/31/2021, 7:43 PMthx!
Elon
03/31/2021, 7:43 PMone question - does that have a significant performance impact? i.e. does it regenerate the segments?
x
Xiang Fu
03/31/2021, 7:58 PM
yes, it will drop current segment and try to recreate and re-consume it
e
Elon
03/31/2021, 7:59 PMOh, so older segments should not be reloaded, i.e. if the kafka topic no longer contains the old data, right?
x
Xiang Fu
03/31/2021, 8:04 PM
you can specify the segment name in that call to reload
👍 1
e
Elon
03/31/2021, 8:04 PMsg, thanks @Xiang Fu!
x
Xiang Fu
03/31/2021, 8:04 PM
if reloading the table, it will touch all segments
e
Elon
03/31/2021, 8:04 PMoh, so we could lose data if the kafka topic no longer contains the offsets from that segment, right?
x
Xiang Fu
03/31/2021, 8:20 PM
yes, if your current segment is very large
👍 1
Xiang Fu
03/31/2021, 8:20 PM
the same issue will happen when you restart the server though
💡 1
e
Elon
03/31/2021, 8:21 PMthis is all really good to know, thanks! It looks like restarting the servers worked for us.
Elon
03/31/2021, 8:21 PMMuch appreciated @Xiang Fu! you helped us again:)