Yunze Xu
I just double checked the issue and it looks like a bug with protobuf: ``` apache/pulsar-client-cpp#1 0x0000ffffb1a58d2c in pulsar::proto::KeyValue::_InternalSerialize(unsigned char*, google::protobuf::io::EpsCopyOutputStream*) const () from /usr/local/lib/python3.10/dist-packages/pulsar_client.libs/libpulsar-8dc83e12.so apache/pulsar-client-cpp#2 0x0000ffffb1cc50e0 in google::protobuf::internal::WireFormatLite::InternalWriteMessage(int, google::protobuf::MessageLite...
Oh I realized where the issue is, there is a bug with the C++ client when batching is disabled. However, the Python client somehow disables batching by default. You can...
> That's the whole point of the managedLedgerMaxReadsInFlightSizeInMB solution that it adds a limit. Just a simple question. When should users modify this config?
> In many cases, it will be useful if the value can be modified if a user faces an issue that could be solved by tuning the limit. The key...
> @merlimat The thread switching was added in PR https://github.com/apache/pulsar/pull/9039, already in December 2020. @merlimat @lhotari to correct it, this is the very early behavior introduced in https://github.com/apache/pulsar/pull/1521. This PR...
Let me close this PR first. > An extra context switch for each entry is costly It's true, but we still need a benchmark. I ran a simple test by...
Out of the scope of this PR, I think it would be better to add a method to `ManagedLedger` for it. Because the customized ML implementation (e.g. StreamNative's Ursa engine)...
Could it break compatibility for old clients?