The IPAM pool analyzer checks that utilization of the pod IP subnet is
less than 85%. For example, if using 10.32.0.0/12, this analyzer will
warn if 3,482 IPs are currently allocated to pods.
The pending allocation analyzer checks that the IPAM status in the
report has no items for the PendingAllocates field. This indicates the
IPAM service is not ready according to the code in the weave status
template
e3712152d2/prog/weaver/http.go (L186).
The weave connections analyzer checks that all connections to remote
peers are in the established state. The state will be "pending" if UDP
is blocked between nodes and will be "failed" if the weave pod on the
remote node is in a crash loop. To force a pending state for testing,
run the commands `iptables -A INPUT -p udp --dport 6784 -j REJECT` and
`iptables -A INPUT -p udp --dport 6783 -j REJECT` on a peer.
The weave connections analyzer also checks that all connections are
using the fastdp protocol. A commopn issue seen in the field on
CentOS/RHEL 7 is that some sides of a connection are using fastdp and
other sides have fallen back to sleeve. Set the WEAVE_NO_FASTDP env var
on the weave daemonset to "true" to test this analyzer.
Fixed bug caused by host preflights not handling empty when clauses, this cropped up because we now handle multiple host preflight results. Also expanded test coverage and added integration test script.
Add a background IOPS feature to the filesystem performance collector
that specifies separate read and write background IOPS to perform while
measuring latency. This allows for better assessment of whether etcd
will be stable when running alongside other workloads on the same
cluster.
Also add templating to the outcome message of the filesystem performance
analyzers to allow printing individual latency percentiles or the entire
table.
Remove the random IOPS benchmark since it was attempting to perform
unaligned direct I/O.
The random IOPS benchmark attempts to replicate the results of this
fio command:
fio --ioengine=psync --direct=1 --bs=4k --size=1Gi --readwrite=randread --serialize_overlap=1
Across three tests the fio command reported 1877 IOPS and the preflight
1822 IOPS with the same block and file size.
The included example found P99 latency of 2.6ms.
Fio reported P99 latency of 2.5ms with this command:
fio --rw=write --ioengine=sync --fdatasync=1 --directory=/var/lib/etcd
--size=220m --bs=2300
