[15:23:02] <Krinkle>	 Seeing a bunch of these in Logstash:
[15:23:04] <Krinkle>	 "Could not enqueue jobs: Unable to deliver all events: 400: 0 out of 1 events were accepted."
[15:23:23] <Krinkle>	 Suggests either Ciruss is sending bad jobs, or EventBus is not accepting them correctly? 
[15:24:20] <_joe_>	 Krinkle: I guess erik or david are your best bet
[15:24:25] <_joe_>	 but they're not here
[15:24:38] <Krinkle>	 mobrovac: ^
[15:25:11] <Krinkle>	 sample at https://logstash.wikimedia.org/app/kibana#/doc/logstash-*/logstash-2019.08.20/mediawiki?id=AWyvnjn-ZKA7Rpirr5FN&_g=h@44136fa
[15:29:22] <mobrovac>	 it does seem like cirrus jobs failing to be sent
[15:29:49] <mobrovac>	 Pchelolo: ^
[15:30:22] <Pchelolo>	 one sec
[15:30:42] <mobrovac>	 the wave has passed though
[15:30:50] <Pchelolo>	 mobrovac: MessageSizeTooLargeError: MESSAGE_SIZE_TOO_LARGE
[15:30:58] <mobrovac>	 ah message too large
[15:30:59] <Pchelolo>	 the jobs are over 15 megs
[15:31:00] <mobrovac>	 yup, just seen it
[15:32:23] <Pchelolo>	 we can file a ticket, I know erik is looking at partitioning the cirrys jobs per cluster, maybe he can try again to finally make them smaller as well.. 
[15:32:39] <Pchelolo>	 but bumbing up the message size even more is definitely not a solution
[15:32:57] <mobrovac>	 no no, agreed, we shouldn't do that