This makes it possible to run an etcd node for testing and development
without placing lots of load on the file system.

Fixes etcd-io#11930.

Signed-off-by: David Crawshaw <[email protected]>

…-#11946-release-3.4

Automated cherry pick of etcd-io#11946

modify file util.

Handle negative value input which currently gives a runtime error.

Handle the related panic and clarify doc.

When an etcd instance attempts to perform service discovery, if a
cluster size with negative value  is provided, the etcd instance
will panic without recovery because of

…ord user

…-#11841-upstream-release-3.4

Automated cherry pick of etcd-io#11841

…-#11830-upstream-release-3.4

Automated cherry pick of etcd-io#11830

…-#11818-upstream-release-3.4

Automated cherry pick of etcd-io#11818

…-#11796-upstream-release-3.4

Automated cherry pick of etcd-io#11796

…-#11608-upstream-release-3.4

Automated cherry pick of etcd-io#11608

…-#11845-upstream-release-3.4

Automated cherry pick of etcd-io#11845

…-#11795-upstream-release-3.4

Automated cherry pick of etcd-io#11795

…-#11807-upstream-release-3.4

Automated cherry pick of etcd-io#11807

…-#11787-upstream-release-3.4

Automated cherry pick of etcd-io#11787

wal: add slice bound checks in decoder

CHANGELOG-3.5: add wal slice bound check
CHANGELOG-3.5: add "decodeRecord"

Signed-off-by: Gyuho Lee <[email protected]>

Currently, watch cancel requests are only sent to the server after a
message comes through on a watch where the client has cancelled. This
means that cancelled watches that don't receive any new messages are
never cancelled; they persist for the lifetime of the client stream.
This has negative connotations for locking applications where a watch
may observe a key which might never change again after cancellation,
leading to many accumulating watches on the server.

By cancelling proactively, in most cases we simply move the cancel
request to happen earlier, and additionally we solve the case where the
cancel request would never be sent.

Fixes etcd-io#9416
Heavy inspiration drawn from the solutions proposed there.

…r will check File Descriptor Requirements ,and count all fds in used. And recorded some logs when in used >= limit/5*4. Just recorded message. If fds was more than 10K,It's low performance due to FDUsage() works. So need to increase it.

see etcd-io#11969 for more detail.

…-#11798-upstream-release-3.4

Automated cherry pick of etcd-io#11798

…of-#12066-origin-release-3.4

Automated cherry pick of etcd-io#12066

…-#12060-upstream-release-3.4

Automated cherry pick of etcd-io#12060

…#11986-upstream-release-3.4

Automated cherry pick of etcd-io#11986

…of-#11850-origin-release-3.4

Automated cherry pick of etcd-io#11850

…-#11793-upstream-release-3.4

Automated cherry pick of etcd-io#11793

…ithout username or username was empty.

Fiexs etcd-io#12004 .

…be enough. Sometimes the consumption speed is more than the production speed.

See etcd-io#11906 for more detail.

see etcd-io#11978 for more detail.

…-#11945-upstream-release-3.4

Automated cherry pick of etcd-io#11945

…#11980-upstream-release-3.4

Automated cherry pick of etcd-io#11980

…#12005-upstream-release-3.4

Automated cherry pick of etcd-io#12005

…-#12092-upstream-release-3.4

Automated cherry pick of etcd-io#12092

…of-#12100-origin-release-3.4

Automated cherry pick of etcd-io#12100

…of-#11997-origin-release-3.4

Automated cherry pick of etcd-io#11997

…#11987-upstream-release-3.4

Automated cherry pick of etcd-io#11987

os.MkdirAll creates directory before umask so make sure that a desired
permission is set after creating a directory with MkdirAll. Use the
existing TouchDirAll function which checks for permission if dir is already
exist and when create a new dir.

…-#12018-upstream-release-3.4

Automated cherry pick of etcd-io#12018

…-#12012-upstream-release-3.4

Automated cherry pick of etcd-io#12012

…e-protobuf-field-type-from-int-to-int64

etcdserver: change protobuf field type from int to int64 (etcd-io#12000)

Signed-off-by: Gyuho Lee <[email protected]>

Signed-off-by: Gyuho Lee <[email protected]>

cr https://code.amazon.com/reviews/CR-29712724

No need sort when we just want the counts.

Signed-off-by: Gyuho Lee <[email protected]>

Similar counts are exposed via Prometheus.
This adds the one that are perceived by etcd server.

e.g.

os_fd_limit 120000
os_fd_used 14
process_cpu_seconds_total 0.31
process_max_fds 120000
process_open_fds 17

Signed-off-by: Gyuho Lee <[email protected]>

…f-#11452-etcd-io#12187-upstream-release-3.4

Automated cherry pick of etcd-io#11452 etcd-io#12187 on release 3.4

Signed-off-by: Gyuho Lee <[email protected]>

Fix bugs introduced in commit c60dabf

…entURLs"

This reverts commit 0372cfc.

Signed-off-by: Gyuho Lee <[email protected]>

*: add plog logging to the backport of PR12216

[3.4] etcdserver: Avoid panics logging slow v2 requests in integration tests

Signed-off-by: Gyuho Lee <[email protected]>

Signed-off-by: Sam Batschelet <[email protected]>

…-of-#12243-upstream-release-3.4

Automated cherry pick of etcd-io#12243 on release 3.4

Provide warning and doc instead of enforcing file permission.

…-#12242-upstream-release-3.4

Automated cherry pick of etcd-io#12242

Signed-off-by: Gyuho Lee <[email protected]>

…f-#12271-upstream-release-3.4

Automated cherry pick of etcd-io#12271 on release 3.4

To fix a panic that happens when trying to get ids of etcd members in
force new cluster mode, the issue happen if the cluster previously had
etcd learner nodes added to the cluster

Fixes etcd-io#12285

[Backport 3.4] etcdserver: add ConfChangeAddLearnerNode to the list of config changes

fixes: etcd-io#11954

…#12264-upstream-release-3.4

Automated cherry pick of etcd-io#12264

To prevent arbitrary command invocations.

Signed-off-by: Gyuho Lee <[email protected]>

Signed-off-by: Gyuho Lee <[email protected]>

Signed-off-by: Gyuho Lee <[email protected]>

Signed-off-by: Sam Batschelet <[email protected]>

Signed-off-by: Sam Batschelet <[email protected]>

Correcting a typo from 2f10964 (version: openshift-v4.0, 2018-11-29).

…uild.

Signed-off-by: Sam Batschelet <[email protected]>

Signed-off-by: Sam Batschelet <[email protected]>

Signed-off-by: Sam Batschelet <[email protected]>

make build performs a sanity test on the binary image which causes problems for unsupport arch. Because we run full CI tests against the image this check is not nessisary and will allow images to be build regardless of arch.

Signed-off-by: Sam Batschelet <[email protected]>

Signed-off-by: Sam Batschelet <[email protected]>

…xists, go ahead and start.