Forum OpenACS Q&A: Re: OpenACS clustering setup and how it relates to xotcl-core.

Thanks Gustaf,

You were right the avgwaittime on pool1 was high

pool1:	statements 1.1M gethandles 8.6K handles 15 connected 0 used 15 waittime 14693.460994 sqltime 1724.604281 avgwaittime 1.7s avgsqltime 1.5ms

The pool1 connections were still at 15. I increased them to 70 and brought our maxthreads down to 60.

Here is my test with this configuration

# 48 cores   200 Gig RAM
# maxconnections 2000
# maxthreads 60
# minthreads 60
# connsperthread 10000
# highwatermark 100
# compressenable off
# rejectoverrun true
# image pool 6/6
# ns_section ns/db/pool/pool1
ns_param connections 70

150 VUs
Connection Threads: min 60 max 60 current 60 idle 33 stopping 0 waiting 0
Request Handling: requests 2.4K, queued 0 (0.00%), spooled 2K (64.39%)
Request Timing: avg queue time 167.94µs, avg filter time 4.87ms, avg run time 231.84ms avg trace time 996.02µs

pool1: statements 288.5K gethandles 2.6K handles 70 connected 5 used 34 waittime 0.477450 sqltime 447.052551 avgwaittime 186.3µs avgsqltime 1.5ms

300 VUs
Connection Threads: min 60 max 60 current 60 idle 18 stopping 0 waiting 0
Request Handling: requests 7.4K, queued 2K (21.37%), spooled 5K (64.86%)
Request Timing: avg queue time 87.8ms, avg filter time 8.85ms, avg run time 333.63ms avg trace time 3.23ms

pool1: statements 923.4K gethandles 7.6K handles 70 connected 17 used 61 waittime 0.998778 sqltime 1997.236181 avgwaittime 132.1µs avgsqltime 2.2ms

150 VUs
Connection Threads: min 60 max 60 current 60 idle 7 stopping 0 waiting 0
Request Handling: requests 11.3K, queued 3K (29.84%), spooled 7K (66.01%)
Request Timing: avg queue time 324.46ms, avg filter time 9.84ms, avg run time 416.1ms avg trace time 3.87ms

pool1: statements 1.5M gethandles 11.5K handles 70 connected 7 used 62 waittime 1.052769 sqltime 3914.273814 avgwaittime 91.8µs avgsqltime 2.5ms

70 VUs
Connection Threads: min 60 max 60 current 60 idle 57 stopping 0 waiting 0
Request Handling: requests 13K, queued 3K (25.79%), spooled 9K (66.52%)
Request Timing: avg queue time 280.5ms, avg filter time 9.11ms, avg run time 414.92ms avg trace time 3.51ms

pool1: statements 1.8M gethandles 13.2K handles 70 connected 57 used 62 waittime 1.054629 sqltime 4429.716496 avgwaittime 79.6µs avgsqltime 2.4ms

K6 Summary
running (4m09.2s), 000/300 VUs, 2225 complete and 0 interrupted iterations
default ✓ [======================================] 000/300 VUs 4m0s

✗ status is 200
↳ 99% — ✓ 8889 / ✗ 1
✗ page succeeded
↳ 99% — ✓ 8886 / ✗ 4

checks.........................: 99.97% ✓ 17775 ✗ 5
data_received..................: 2.1 GB 8.3 MB/s
data_sent......................: 11 MB 44 kB/s
http_req_blocked...............: avg=2.35ms min=1.27µs med=4.83µs max=1.85s p(90)=4.9ms p(95)=7.77ms
http_req_connecting............: avg=762.21µs min=0s med=0s max=1.84s p(90)=340.97µs p(95)=392.74µs
✓ http_req_duration..............: avg=699.68ms min=4.73ms med=43.58ms max=5.8s p(90)=2.63s p(95)=3.23s
{ expected_response:true }...: avg=699.74ms min=4.73ms med=43.59ms max=5.8s p(90)=2.63s p(95)=3.23s
http_req_failed................: 0.00% ✓ 1 ✗ 13342
http_req_receiving.............: avg=927.2µs min=42.66µs med=167.63µs max=1.63s p(90)=1.68ms p(95)=2.29ms
http_req_sending...............: avg=32.23µs min=7.95µs med=28.46µs max=472.34µs p(90)=58.33µs p(95)=66.47µs
http_req_tls_handshaking.......: avg=1.56ms min=0s med=0s max=37.52ms p(90)=4.48ms p(95)=7.27ms
http_req_waiting...............: avg=698.72ms min=4.45ms med=43.41ms max=5.8s p(90)=2.63s p(95)=3.23s
http_reqs......................: 13343 53.54756/s
iteration_duration.............: avg=14.18s min=1.74µs med=12.87s max=22.24s p(90)=18.83s p(95)=19.99s
iterations.....................: 2225 8.929275/s
vus............................: 1 min=1 max=300
vus_max........................: 300 min=300 max=300

These numbers look a lot better! If you see anything that seems amiss let me know. We appreciate all of your help and expertise.

Sincerely,
Marty