[10:17:56] <gehel>	 dcausse: and since we're talking about code reviews: https://gerrit.wikimedia.org/r/c/wikimedia-event-utilities/+/1090885
[10:18:29] <dcausse>	 gehel: I did not forget about this one I promess :)
[10:31:56] <dcausse>	 errand+lunch
[10:40:24] <gehel>	 :)
[15:19:43] <pfischer>	 Hi, I started working with jupyter (via hub) to query iceberg tables via spark. If want to export the contents of a dataframe, how do I know where df.write.csv(‘some/path’) ends up (physycially)? It’s not the stat machine I use to tunnel into jupyter hub, at least I can’t find that file there.
[15:27:20] <pfischer>	 I guess I have to write to HDFS
[15:35:12] <dcausse>	 pfischer: yes better to write to hdfs as a single partition and then use hdfs dfs -text hdfs:///patch_to_file.csv
[15:39:08] <pfischer>	 dcausse: Thanks! If I use an absolute path (hdfs:///analytics-hadoop/users/pfischer…) for writing, I do get an error: Permission Denied. Using hdfs://analytics-hadoop/users/pfischer… works, but hdfs dfs -ls analytics-hadoop/users/pfischer says that no such directory exists.
[15:40:47] <dcausse>	 pfischer: without the hdfs:// scheme it should "hdfs dfs -ls /user/pfischer" I think
[15:41:31] <pfischer>	 Ah, TIL… Thank you once more!
[15:42:14] <dcausse>	 AFAIK it can be: hdfs://analytics-hadoop/user/pfischer, hdfs:///user/pfischer or /user/pfischer
[17:52:52] <JJMC89>	 !issync
[17:52:53] <ircservserv-wm>	 Syncing #wikimedia-search (requested by JJMC89)
[17:52:54] <ircservserv-wm>	 Set /cs flags #wikimedia-search wmopbot +t
[17:52:56] <ircservserv-wm>	 Set /cs flags #wikimedia-search Az1568 -AFRefiorstv
[18:28:13] <dcausse>	 dinner