kotovalexarian-likes-github/moby--moby
1
0
Fork 0
mirror of https://github.com/moby/moby.git synced 2022-11-09 12:21:53 -05:00
Code Releases Activity
moby--moby/daemon/discovery/discovery_test.go

112 lines
3.1 KiB
Go
Raw Normal View History

Extract daemon configuration and discovery to their own package This also moves some cli specific in `cmd/dockerd` as it does not really belong to the `daemon/config` package. Signed-off-by: Vincent Demeester <vincent@sbr.pm>
2017-01-23 06:23:07 -05:00
package discovery
Make discovery ttl and heartbeat configurable Docker daemon uses kv-store as the host-discovery backend. Discovery module tracks the liveness of a node through a simple keepalive mechanism. The keepalive mechanism depends on every node performing heartbeat by registering itself with the discovery module (via KV-Store Put operation). And for every Put operation, the discovery module in all other nodes will receive a Watch notification. That keeps the node alive. Any node that fails to register itself within the TTL timer is considered dead and removed from the discovery database. The default timer (heartbeat = 20 seconds & ttl = 60 seconds) works fine for small clusters. But for large clusters, these default timers are extremely aggressive and that causes high CPU & most of the processing is spent managing the node discovery and that impacts normal daemon operation. Hence we need a way to make the discovery ttl and heartbeat configurable. As the cluster size grows, the user can change these timers to make sure the daemon scales. Signed-off-by: Madhu Venugopal <madhu@docker.com>
2015-11-11 19:18:06 -05:00
import (
The variable heartbeat might be 0 Signed-off-by: Yanqiang Miao <miao.yanqiang@zte.com.cn> perfect the test case for 'discoveryOpts' Signed-off-by: Yanqiang Miao <miao.yanqiang@zte.com.cn>
2017-03-11 03:03:22 -05:00
"fmt"
Make discovery ttl and heartbeat configurable Docker daemon uses kv-store as the host-discovery backend. Discovery module tracks the liveness of a node through a simple keepalive mechanism. The keepalive mechanism depends on every node performing heartbeat by registering itself with the discovery module (via KV-Store Put operation). And for every Put operation, the discovery module in all other nodes will receive a Watch notification. That keeps the node alive. Any node that fails to register itself within the TTL timer is considered dead and removed from the discovery database. The default timer (heartbeat = 20 seconds & ttl = 60 seconds) works fine for small clusters. But for large clusters, these default timers are extremely aggressive and that causes high CPU & most of the processing is spent managing the node discovery and that impacts normal daemon operation. Hence we need a way to make the discovery ttl and heartbeat configurable. As the cluster size grows, the user can change these timers to make sure the daemon scales. Signed-off-by: Madhu Venugopal <madhu@docker.com>
2015-11-11 19:18:06 -05:00
"testing"
"time"
)
func TestDiscoveryOpts(t *testing.T) {
clusterOpts := map[string]string{"discovery.heartbeat": "10", "discovery.ttl": "5"}
heartbeat, ttl, err := discoveryOpts(clusterOpts)
if err == nil {
use t.Fatal() to output the err message where the values used for formatting text does not appear to contain a placeholder Signed-off-by: Helen Xie <chenjg@harmonycloud.cn>
2017-02-21 03:53:29 -05:00
t.Fatal("discovery.ttl < discovery.heartbeat must fail")
Make discovery ttl and heartbeat configurable Docker daemon uses kv-store as the host-discovery backend. Discovery module tracks the liveness of a node through a simple keepalive mechanism. The keepalive mechanism depends on every node performing heartbeat by registering itself with the discovery module (via KV-Store Put operation). And for every Put operation, the discovery module in all other nodes will receive a Watch notification. That keeps the node alive. Any node that fails to register itself within the TTL timer is considered dead and removed from the discovery database. The default timer (heartbeat = 20 seconds & ttl = 60 seconds) works fine for small clusters. But for large clusters, these default timers are extremely aggressive and that causes high CPU & most of the processing is spent managing the node discovery and that impacts normal daemon operation. Hence we need a way to make the discovery ttl and heartbeat configurable. As the cluster size grows, the user can change these timers to make sure the daemon scales. Signed-off-by: Madhu Venugopal <madhu@docker.com>
2015-11-11 19:18:06 -05:00
}
clusterOpts = map[string]string{"discovery.heartbeat": "10", "discovery.ttl": "10"}
heartbeat, ttl, err = discoveryOpts(clusterOpts)
if err == nil {
use t.Fatal() to output the err message where the values used for formatting text does not appear to contain a placeholder Signed-off-by: Helen Xie <chenjg@harmonycloud.cn>
2017-02-21 03:53:29 -05:00
t.Fatal("discovery.ttl == discovery.heartbeat must fail")
Make discovery ttl and heartbeat configurable Docker daemon uses kv-store as the host-discovery backend. Discovery module tracks the liveness of a node through a simple keepalive mechanism. The keepalive mechanism depends on every node performing heartbeat by registering itself with the discovery module (via KV-Store Put operation). And for every Put operation, the discovery module in all other nodes will receive a Watch notification. That keeps the node alive. Any node that fails to register itself within the TTL timer is considered dead and removed from the discovery database. The default timer (heartbeat = 20 seconds & ttl = 60 seconds) works fine for small clusters. But for large clusters, these default timers are extremely aggressive and that causes high CPU & most of the processing is spent managing the node discovery and that impacts normal daemon operation. Hence we need a way to make the discovery ttl and heartbeat configurable. As the cluster size grows, the user can change these timers to make sure the daemon scales. Signed-off-by: Madhu Venugopal <madhu@docker.com>
2015-11-11 19:18:06 -05:00
}
validate heartbeat in discovery cluster opts Signed-off-by: allencloud <allen.sun@daocloud.io>
2016-08-31 13:07:05 -04:00
clusterOpts = map[string]string{"discovery.heartbeat": "-10", "discovery.ttl": "10"}
heartbeat, ttl, err = discoveryOpts(clusterOpts)
if err == nil {
use t.Fatal() to output the err message where the values used for formatting text does not appear to contain a placeholder Signed-off-by: Helen Xie <chenjg@harmonycloud.cn>
2017-02-21 03:53:29 -05:00
t.Fatal("negative discovery.heartbeat must fail")
validate heartbeat in discovery cluster opts Signed-off-by: allencloud <allen.sun@daocloud.io>
2016-08-31 13:07:05 -04:00
}
clusterOpts = map[string]string{"discovery.heartbeat": "10", "discovery.ttl": "-10"}
heartbeat, ttl, err = discoveryOpts(clusterOpts)
if err == nil {
use t.Fatal() to output the err message where the values used for formatting text does not appear to contain a placeholder Signed-off-by: Helen Xie <chenjg@harmonycloud.cn>
2017-02-21 03:53:29 -05:00
t.Fatal("negative discovery.ttl must fail")
validate heartbeat in discovery cluster opts Signed-off-by: allencloud <allen.sun@daocloud.io>
2016-08-31 13:07:05 -04:00
}
Make discovery ttl and heartbeat configurable Docker daemon uses kv-store as the host-discovery backend. Discovery module tracks the liveness of a node through a simple keepalive mechanism. The keepalive mechanism depends on every node performing heartbeat by registering itself with the discovery module (via KV-Store Put operation). And for every Put operation, the discovery module in all other nodes will receive a Watch notification. That keeps the node alive. Any node that fails to register itself within the TTL timer is considered dead and removed from the discovery database. The default timer (heartbeat = 20 seconds & ttl = 60 seconds) works fine for small clusters. But for large clusters, these default timers are extremely aggressive and that causes high CPU & most of the processing is spent managing the node discovery and that impacts normal daemon operation. Hence we need a way to make the discovery ttl and heartbeat configurable. As the cluster size grows, the user can change these timers to make sure the daemon scales. Signed-off-by: Madhu Venugopal <madhu@docker.com>
2015-11-11 19:18:06 -05:00
clusterOpts = map[string]string{"discovery.heartbeat": "invalid"}
heartbeat, ttl, err = discoveryOpts(clusterOpts)
if err == nil {
use t.Fatal() to output the err message where the values used for formatting text does not appear to contain a placeholder Signed-off-by: Helen Xie <chenjg@harmonycloud.cn>
2017-02-21 03:53:29 -05:00
t.Fatal("invalid discovery.heartbeat must fail")
Make discovery ttl and heartbeat configurable Docker daemon uses kv-store as the host-discovery backend. Discovery module tracks the liveness of a node through a simple keepalive mechanism. The keepalive mechanism depends on every node performing heartbeat by registering itself with the discovery module (via KV-Store Put operation). And for every Put operation, the discovery module in all other nodes will receive a Watch notification. That keeps the node alive. Any node that fails to register itself within the TTL timer is considered dead and removed from the discovery database. The default timer (heartbeat = 20 seconds & ttl = 60 seconds) works fine for small clusters. But for large clusters, these default timers are extremely aggressive and that causes high CPU & most of the processing is spent managing the node discovery and that impacts normal daemon operation. Hence we need a way to make the discovery ttl and heartbeat configurable. As the cluster size grows, the user can change these timers to make sure the daemon scales. Signed-off-by: Madhu Venugopal <madhu@docker.com>
2015-11-11 19:18:06 -05:00
}
clusterOpts = map[string]string{"discovery.ttl": "invalid"}
heartbeat, ttl, err = discoveryOpts(clusterOpts)
if err == nil {
use t.Fatal() to output the err message where the values used for formatting text does not appear to contain a placeholder Signed-off-by: Helen Xie <chenjg@harmonycloud.cn>
2017-02-21 03:53:29 -05:00
t.Fatal("invalid discovery.ttl must fail")
Make discovery ttl and heartbeat configurable Docker daemon uses kv-store as the host-discovery backend. Discovery module tracks the liveness of a node through a simple keepalive mechanism. The keepalive mechanism depends on every node performing heartbeat by registering itself with the discovery module (via KV-Store Put operation). And for every Put operation, the discovery module in all other nodes will receive a Watch notification. That keeps the node alive. Any node that fails to register itself within the TTL timer is considered dead and removed from the discovery database. The default timer (heartbeat = 20 seconds & ttl = 60 seconds) works fine for small clusters. But for large clusters, these default timers are extremely aggressive and that causes high CPU & most of the processing is spent managing the node discovery and that impacts normal daemon operation. Hence we need a way to make the discovery ttl and heartbeat configurable. As the cluster size grows, the user can change these timers to make sure the daemon scales. Signed-off-by: Madhu Venugopal <madhu@docker.com>
2015-11-11 19:18:06 -05:00
}
clusterOpts = map[string]string{"discovery.heartbeat": "10", "discovery.ttl": "20"}
heartbeat, ttl, err = discoveryOpts(clusterOpts)
if err != nil {
t.Fatal(err)
}
if heartbeat != 10*time.Second {
correct some nits in comment and test files Signed-off-by: allencloud <allen.sun@daocloud.io>
2016-08-29 06:37:14 -04:00
t.Fatalf("Heartbeat - Expected : %v, Actual : %v", 10*time.Second, heartbeat)
Make discovery ttl and heartbeat configurable Docker daemon uses kv-store as the host-discovery backend. Discovery module tracks the liveness of a node through a simple keepalive mechanism. The keepalive mechanism depends on every node performing heartbeat by registering itself with the discovery module (via KV-Store Put operation). And for every Put operation, the discovery module in all other nodes will receive a Watch notification. That keeps the node alive. Any node that fails to register itself within the TTL timer is considered dead and removed from the discovery database. The default timer (heartbeat = 20 seconds & ttl = 60 seconds) works fine for small clusters. But for large clusters, these default timers are extremely aggressive and that causes high CPU & most of the processing is spent managing the node discovery and that impacts normal daemon operation. Hence we need a way to make the discovery ttl and heartbeat configurable. As the cluster size grows, the user can change these timers to make sure the daemon scales. Signed-off-by: Madhu Venugopal <madhu@docker.com>
2015-11-11 19:18:06 -05:00
}
if ttl != 20*time.Second {
t.Fatalf("TTL - Expected : %v, Actual : %v", 20*time.Second, ttl)
}
clusterOpts = map[string]string{"discovery.heartbeat": "10"}
heartbeat, ttl, err = discoveryOpts(clusterOpts)
if err != nil {
t.Fatal(err)
}
if heartbeat != 10*time.Second {
correct some nits in comment and test files Signed-off-by: allencloud <allen.sun@daocloud.io>
2016-08-29 06:37:14 -04:00
t.Fatalf("Heartbeat - Expected : %v, Actual : %v", 10*time.Second, heartbeat)
Make discovery ttl and heartbeat configurable Docker daemon uses kv-store as the host-discovery backend. Discovery module tracks the liveness of a node through a simple keepalive mechanism. The keepalive mechanism depends on every node performing heartbeat by registering itself with the discovery module (via KV-Store Put operation). And for every Put operation, the discovery module in all other nodes will receive a Watch notification. That keeps the node alive. Any node that fails to register itself within the TTL timer is considered dead and removed from the discovery database. The default timer (heartbeat = 20 seconds & ttl = 60 seconds) works fine for small clusters. But for large clusters, these default timers are extremely aggressive and that causes high CPU & most of the processing is spent managing the node discovery and that impacts normal daemon operation. Hence we need a way to make the discovery ttl and heartbeat configurable. As the cluster size grows, the user can change these timers to make sure the daemon scales. Signed-off-by: Madhu Venugopal <madhu@docker.com>
2015-11-11 19:18:06 -05:00
}
expected := 10 * defaultDiscoveryTTLFactor * time.Second
if ttl != expected {
t.Fatalf("TTL - Expected : %v, Actual : %v", expected, ttl)
}
clusterOpts = map[string]string{"discovery.ttl": "30"}
heartbeat, ttl, err = discoveryOpts(clusterOpts)
if err != nil {
t.Fatal(err)
}
if ttl != 30*time.Second {
t.Fatalf("TTL - Expected : %v, Actual : %v", 30*time.Second, ttl)
}
expected = 30 * time.Second / defaultDiscoveryTTLFactor
if heartbeat != expected {
correct some nits in comment and test files Signed-off-by: allencloud <allen.sun@daocloud.io>
2016-08-29 06:37:14 -04:00
t.Fatalf("Heartbeat - Expected : %v, Actual : %v", expected, heartbeat)
Make discovery ttl and heartbeat configurable Docker daemon uses kv-store as the host-discovery backend. Discovery module tracks the liveness of a node through a simple keepalive mechanism. The keepalive mechanism depends on every node performing heartbeat by registering itself with the discovery module (via KV-Store Put operation). And for every Put operation, the discovery module in all other nodes will receive a Watch notification. That keeps the node alive. Any node that fails to register itself within the TTL timer is considered dead and removed from the discovery database. The default timer (heartbeat = 20 seconds & ttl = 60 seconds) works fine for small clusters. But for large clusters, these default timers are extremely aggressive and that causes high CPU & most of the processing is spent managing the node discovery and that impacts normal daemon operation. Hence we need a way to make the discovery ttl and heartbeat configurable. As the cluster size grows, the user can change these timers to make sure the daemon scales. Signed-off-by: Madhu Venugopal <madhu@docker.com>
2015-11-11 19:18:06 -05:00
}
The variable heartbeat might be 0 Signed-off-by: Yanqiang Miao <miao.yanqiang@zte.com.cn> perfect the test case for 'discoveryOpts' Signed-off-by: Yanqiang Miao <miao.yanqiang@zte.com.cn>
2017-03-11 03:03:22 -05:00
discaveryTTL := fmt.Sprintf("%d", defaultDiscoveryTTLFactor-1)
clusterOpts = map[string]string{"discovery.ttl": discaveryTTL}
heartbeat, ttl, err = discoveryOpts(clusterOpts)
if err == nil && heartbeat == 0 {
t.Fatal("discovery.heartbeat must be positive")
}
Make discovery ttl and heartbeat configurable Docker daemon uses kv-store as the host-discovery backend. Discovery module tracks the liveness of a node through a simple keepalive mechanism. The keepalive mechanism depends on every node performing heartbeat by registering itself with the discovery module (via KV-Store Put operation). And for every Put operation, the discovery module in all other nodes will receive a Watch notification. That keeps the node alive. Any node that fails to register itself within the TTL timer is considered dead and removed from the discovery database. The default timer (heartbeat = 20 seconds & ttl = 60 seconds) works fine for small clusters. But for large clusters, these default timers are extremely aggressive and that causes high CPU & most of the processing is spent managing the node discovery and that impacts normal daemon operation. Hence we need a way to make the discovery ttl and heartbeat configurable. As the cluster size grows, the user can change these timers to make sure the daemon scales. Signed-off-by: Madhu Venugopal <madhu@docker.com>
2015-11-11 19:18:06 -05:00
clusterOpts = map[string]string{}
heartbeat, ttl, err = discoveryOpts(clusterOpts)
if err != nil {
t.Fatal(err)
}
if heartbeat != defaultDiscoveryHeartbeat {
correct some nits in comment and test files Signed-off-by: allencloud <allen.sun@daocloud.io>
2016-08-29 06:37:14 -04:00
t.Fatalf("Heartbeat - Expected : %v, Actual : %v", defaultDiscoveryHeartbeat, heartbeat)
Make discovery ttl and heartbeat configurable Docker daemon uses kv-store as the host-discovery backend. Discovery module tracks the liveness of a node through a simple keepalive mechanism. The keepalive mechanism depends on every node performing heartbeat by registering itself with the discovery module (via KV-Store Put operation). And for every Put operation, the discovery module in all other nodes will receive a Watch notification. That keeps the node alive. Any node that fails to register itself within the TTL timer is considered dead and removed from the discovery database. The default timer (heartbeat = 20 seconds & ttl = 60 seconds) works fine for small clusters. But for large clusters, these default timers are extremely aggressive and that causes high CPU & most of the processing is spent managing the node discovery and that impacts normal daemon operation. Hence we need a way to make the discovery ttl and heartbeat configurable. As the cluster size grows, the user can change these timers to make sure the daemon scales. Signed-off-by: Madhu Venugopal <madhu@docker.com>
2015-11-11 19:18:06 -05:00
}
expected = defaultDiscoveryHeartbeat * defaultDiscoveryTTLFactor
if ttl != expected {
t.Fatalf("TTL - Expected : %v, Actual : %v", expected, ttl)
}
}
Copy permalink