many workers threads failed with EAGAIN
David Birdsong
david.birdsong at gmail.com
Mon Aug 31 14:36:51 CEST 2009
varnishlog has a lot of these:
0 Debug - "Create worker thread failed 11 Resource temporarily
unavailable"
sure enough, overflowed and dropped work requests are steadily on the rise
Hitrate ratio: 10 100 1000
Hitrate avg: 0.8584 0.8506 0.8581
1123681 423.11 380.14 Client connections accepted
706252 189.05 238.92 Client requests received
587166 159.04 198.64 Cache hits
6928 3.00 2.34 Cache hits for pass
111419 27.01 37.69 Cache misses
119397 30.01 40.39 Backend conn. success
2 0.00 0.00 Backend conn. not attempted
5744 . . N struct sess_mem
3663 . . N struct sess
95290 . . N struct object
95912 . . N struct objecthead
6 . . N struct vbe_conn
1007 . . N worker threads
1007 0.00 0.34 N worker threads created
12425 4.00 4.20 N worker threads not created
2002 1.00 0.68 N queued work requests
651706 157.04 220.47 N overflowed work requests
460396 262.07 155.75 N dropped work requests
9 . . N backends
16469 . . N LRU nuked objects
0 . . N LRU moved objects
5 0.00 0.00 HTTP header overflows
580405 158.04 196.35 Objects sent with write
674521 186.05 228.19 Total Sessions
706252 189.05 238.92 Total Requests
904 0.00 0.31 Total pipe
6928 3.00 2.34 Total pass
118116 30.01 39.96 Total fetch
189339016 50665.57 64052.44 Total header bytes
49792939291 18055006.14 16844702.06 Total body bytes
143000 47.01 48.38 Session Closed
779 0.00 0.26 Session Pipeline
83 0.00 0.03 Session Read Ahead
609288 155.04 206.12 Session Linger
My arguments to start varnishd:
ulimit -n 500000
# stack size
ulimit -s 256
/usr/local/sbin/varnishd \
-T localhost:6082 \
-f $VCL \
-s malloc,10GB \
-P /var/run/varnish.pid \
-h classic,1000003 \
-p lru_interval=3600 \
-p thread_pool_min=500 \
-p thread_pool_max=2000 \
-p thread_pools=2 \
-p obj_workspace=4096 \
-u apache \
-p listen_depth=4096
More information about the varnish-misc
mailing list