Differences

This shows you the differences between two versions of the page.

--- litespeed_wiki:lslb:ha [2020/06/20 15:50]
Lisa Clarke [How two HA ADC nodes handle cache purge and cache sync?] Edited
+++ litespeed_wiki:lslb:ha [2020/08/12 18:44]
Lisa Clarke [Syncing Cache Between Two HA Nodes] Editing
@@ Line 230: / Line 230: @@
 Try accessing 10.10.31.31 (VIP) from the browser. You will see the backend server page. Disable one node, and you can still see the webpage. Check ADC HA status. The live node will become Master when the other one down.
-===== Syncing Cache Between Multiple HA Nodes =====
+===== Syncing Cache Between Two HA Nodes =====
-Cache replication can be enabled or disabled as desired. To enable cache replication, navigate to **ADC WebAdmin Console > Configuration > HA** and set **Enable Replication for Cached Data** to ''Yes''.
+Cache purge is automatically synced between two HA nodes. When a node receives a purge request, the cached file will not be deleted immediately. It will be regenerated when a new request for that URL is received.
+Further cache replication can be enabled or disabled as desired. To enable cache replication, navigate to **ADC WebAdmin Console > Configuration > HA** and set **Enable Replication for Cached Data** to ''Yes''.
 {{ :litespeed_wiki:lslb:adc-ha-enablecachereplication.png?800 |}}
-When replication is enabled, caches are synced between all HA ADC nodes. When the backend LiteSpeed Cache plugin sends a purge request to the live ADC, that request - a "purge" command, for example - will be synced with the nodes.
+When replication is enabled, caches are synced between the HA ADC nodes. When the backend LiteSpeed Cache plugin sends a cache request to the live ADC, that request will be synced with the nodes.
+**NOTE 1**: Replication happens only between two front end High Availability ADCs, and //not// backend cluster web servers. Avoid caching on the cluster-server-node layer, as cached data within web server nodes will not be purged properly and will no longer be synchronized. In a clustered environment, LSCache data should //only be maintained at the ADC layer//.
-When a node receives a purge request, the cached file will not be deleted immediately. It will be regenerated when a new request for that URL is received.
+**NOTE 2**: Only LSCache data is synced between two HA ADCs. //Configuration is not synced//. Therefore, the ADCs must be manually configured independently to serve the same sites.
-**Please note**: Replication happens only between two front end High Availability ADCs, and //not// backend cluster web servers. Avoid caching on the cluster-server-node layer, as cached data within web server nodes will not be purged properly and will no longer be synchronized. In a clustered environment, LSCache data should //only be maintained at the ADC layer//.
 ===== Troubleshooting =====