Venkatesh J

I am of the opinion that running QAT on NUMA node 0 or 1 should not make a difference in performance.. https://www.intel.com/content/www/us/en/developer/articles/technical/use-intel-quickassist-technology-efficiently-with-numa-awareness.html Having said that, can you give more info...

Apologies for the late reply. Your NUMA layout looks fine to me. As you had said that QAT is slow, can you share the comparison test data ? Since the...

I see that in the specification, AES-CBC gets around 103Gbps for 4KB packet size. https://cdrdv2-public.intel.com/691474/intel-quickassist-adapter-8960-8970.pdf I will check on this and get back to you soon..

We just came across a similar issue reported internally. Quoting from the issue, "On a multi-socket platform, there can be a performance degradation on the remote sockets. This can arise...

Hi @lastpepole , thanks for reporting the issue. We will have someone look into this.

@lastpepole , we have created internal defect to track this. Can you share the nginx.conf ?

We are unable to reproduce the deadlock scenario in our test env. The QAT Engine offload seems fine. Driver: QAT.L.4.23.0-00001 Openssl Version: openssl-3.0.13 Qatengine: v1.5.0 Tengine: master branch Certs and...

Hi @lastpepole , we see "Process nginx exit with orphan rings" error sporadically. But we are unable to reproduce the pthread deadlock scenario.. To further narrow down the problem, please...

Thank you for sharing the requested details. We will update you shortly on the issue reproduction with OpenSSL 1.1.1.

> @venkatesh6911 My machine environment configuration as below: kernel: 5.10 os: centos7u5 qat driver: 4.25.0 qat engine: 1.5.0 openssl: 1.1.1w tengine: tengine-2.4.0 > > we are enable to reproduce the...