Israel Fruchter

Results 813 comments of Israel Fruchter

and the plot thickens a bit, we recent update the ssl version we are using in this test from `ssl.PROTOCOL_TLSv1_2` to `sl.PROTOCOL_TLS_CLIENT` (which mean lastet i.e. v3) and with `ssl.PROTOCOL_TLSv1_2`...

looking at seastar recent changes, this change might be affecting the the closing of connections logic ? scylladb/seastar@7164ae143e0d0493a1c8bd91b45e9c0c87ab5deb

> Is there a driver version where things work, and one where it does not? Again, I cannot repro this at all. `PROTOCOL_TLS_CLIENT` should mean we negotiate to TLSv3, but...

> > It happen with the latest driver version 3.26.3, and latest scylla master > > > > it happens about 1/10 like that: > > > > ``` >...

https://github.com/scylladb/scylla-dtest/pull/3670 is hiding this issue for now, as far as I can see. TODO: run the test with tcpdump, so we working network dump, and failing network dump, to compare

and now with tcpdumps: ``` aws s3 cp s3://scylla-qa-public/fruch/python-driver-issue-261/log_and_tcpdump.tar.gz . ``` the failed dump: ``` logs/failure/ccm_node.crl logs/failure/ccm_node.cer logs/failure/ccm_node.pem logs/failure/ccm_node.key logs/failure/tcpdump__127.0.29.1_19042.pcap logs/failure/tcpdump__127.0.29.1_9042.pcap ``` two example of the test passing ``` logs/1697665964613_limits_test.py::TestLimits::test_request_too_large[SSL-8-10]/...

I figured I've captured only one direction of the communication, so here's one that capture both directions (similar directory structure): ``` aws s3 cp s3://scylla-qa-public/fruch/python-driver-issue-261/log_and_tcpdump_2_direction.tar.gz ```

can also be download by https: ``` wget https://scylla-qa-public.s3.amazonaws.com/fruch/python-driver-issue-261/log_and_tcpdump_2_direction.tar.gz ```

seem like there more test thats are affected by this default: NettyOptionsTest.should_invoke_netty_options_hooks_single_node

> So let's revert the offending commit? For reference, this is the change: https://github.com/scylladb/scylla-ccm/pull/436 As I stated in that PR, change shouldn't be that complex (python-driver is working fine with...