Cluster: Remote remove received....

As we transition to the new Hyland Community, some post links may be temporarily disrupted as we finalize redirects. We apologize for the inconvenience and are working to resolve this as quickly as possible. Details.

marco_altieri_7

Ciao a tutti,

ho sottoposto lo stesso problema nel forum americano, ma non ho ricevuto risposta. Spero di essere più fortunato qui.

Ho configurato un cluster di alfresco (versione 3.2r2) con due nodi. La configurazione è assolutamente standard: il db ed il filesystem sono condivisi e le ehcache sono in cluster (ho utilizzato come base di partenza ehcache-custom.xml.sample.cluster). Ho provato anche a non utilizzare jgroups, ma non è cambiato nulla.

Non appena i server cominciano a "parlarsi" nel file di log di entrambi compaiono dei messaggi del tipo:

18:44:15,745  DEBUG [ehcache.distribution.RMICachePeer] Remote remove received for key: workspace://SpacesStore/a9a9f477-5c57-11dc-ad6c-5136d620963c
18:44:15,747  DEBUG [ehcache.distribution.RMICachePeer] Remote remove received for key: workspace://SpacesStore
18:44:15,750  DEBUG [ehcache.distribution.RMICachePeer] Remote remove received for key: workspace://SpacesStore/c317f789-5c57-11dc-ad6c-5136d620963c
18:44:15,753  DEBUG [ehcache.distribution.RMICachePeer] Remote remove received for key: system://system/5f5a740f-2589-4629-bebe-2b63e7176b93
18:44:15,754  DEBUG [ehcache.distribution.RMICachePeer] Remote remove received for key:
18:44:15,760  DEBUG [ehcache.distribution.RMICachePeer] Remote remove received for key: http://www.alfresco.org/model/system/1.0
18:44:15,762  DEBUG [ehcache.distribution.RMICachePeer] Remote remove received for key: 2
18:44:15,762  DEBUG [ehcache.distribution.RMICachePeer] Remote remove received for key: http://www.alfresco.org/model/application/1.0
18:44:15,763  DEBUG [ehcache.distribution.RMICachePeer] Remote remove received for key: 14
18:44:15,764  DEBUG [ehcache.distribution.RMICachePeer] Remote remove received for key: http://www.alfresco.org/model/rule/1.0
18:44:15,765  DEBUG [ehcache.distribution.RMICachePeer] Remote remove received for key: 16
18:44:15,767  DEBUG [ehcache.distribution.RMICachePeer] Remote remove received for key: en_US_
18:44:15,774  DEBUG [ehcache.distribution.RMICachePeer] Remote remove received for key: 4
18:44:15,775  DEBUG [ehcache.distribution.RMICachePeer] Remote remove received for key: text/plain
18:44:15,776  DEBUG [ehcache.distribution.RMICachePeer] Remote remove received for key: 2
18:44:15,777  DEBUG [ehcache.distribution.RMICachePeer] Remote remove received for key: utf8
18:44:15,778  DEBUG [ehcache.distribution.RMICachePeer] Remote remove received for key: 2
18:44:15,779  DEBUG [ehcache.distribution.RMICachePeer] Remote remove received for key: 4
18:44:15,781  DEBUG [ehcache.distribution.RMICachePeer] Remote remove received for key: 24
18:44:15,782  DEBUG [ehcache.distribution.RMICachePeer] Remote remove received for key: {http://www.alfresco.org/model/system/1.0}descriptor
18:44:15,784  DEBUG [ehcache.distribution.RMICachePeer] Remote remove received for key: {http://www.alfresco.org/model/system/1.0}children
18:44:15,786  DEBUG [ehcache.distribution.RMICachePeer] Remote remove received for key: 26
18:44:15,787  DEBUG [ehcache.distribution.RMICachePeer] Remote remove received for key: {http://www.alfresco.org/model/system/1.0}versionSchema
18:44:15,789  DEBUG [ehcache.distribution.RMICachePeer] Remote remove received for key: {http://www.alfresco.org/model/system/1.0}store_root
18:44:15,790  DEBUG [ehcache.distribution.RMICachePeer] Remote remove received for key: {http://www.alfresco.org/model/content/1.0}folder
18:44:15,792  DEBUG [ehcache.distribution.RMICachePeer] Remote remove received for key: 8
18:44:15,793  DEBUG [ehcache.distribution.RMICachePeer] Remote remove received for key: {http://www.alfresco.org/model/content/1.0}contains
18:44:15,794  DEBUG [ehcache.distribution.RMICachePeer] Remote remove received for key: 50
18:44:15,795  DEBUG [ehcache.distribution.RMICachePeer] Remote remove received for key: 68
18:44:15,798  DEBUG [ehcache.distribution.RMICachePeer] Remote remove received for key: {http://www.alfresco.org/model/rule/1.0}ruleFolder
18:44:15,799  DEBUG [ehcache.distribution.RMICachePeer] Remote remove received for key: 122
18:44:15,801  DEBUG [ehcache.distribution.RMICachePeer] Remote remove received for key: {http://www.alfresco.org/model/content/1.0}systemfolder
18:44:15,802  DEBUG [ehcache.distribution.RMICachePeer] Remote remove received for key: 124
18:44:15,803  DEBUG [ehcache.distribution.RMICachePeer] Remote remove received for key: 28
18:44:15,804  DEBUG [ehcache.distribution.RMICachePeer] Remote remove received for key: {http://www.alfresco.org/model/system/1.0}versionMajor
18:44:17,306  DEBUG [ehcache.distribution.RMICachePeer] Remote remove received for key: 30
18:44:17,312  DEBUG [ehcache.distribution.RMICachePeer] Remote remove received for key: {http://www.alfresco.org/model/system/1.0}versionLabel
18:44:17,313  DEBUG [ehcache.distribution.RMICachePeer] Remote remove received for key: 32
18:44:17,314  DEBUG [ehcache.distribution.RMICachePeer] Remote remove received for key: {http://www.alfresco.org/model/system/1.0}versionRevision
18:44:17,315  DEBUG [ehcache.distribution.RMICachePeer] Remote remove received for key: 34
‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍

Le "key" presenti nei messaggi di log spesso si ripetono. Il sistema sembra sostanzialmente bloccato in un loop infinito.
Se spengo e faccio ripartire solo uno dei due server, sembra tutto ok e nessun nodo è stato realmente cancellato.

Temo che  siano state effettuate modifiche su entrambi i nodi quando il cluster non era ancora pronto. Questo potrebbe aver creato dei disallineamenti che adesso non sono più gestibili?
Ho visto che esiste una tabella alf_transaction in cui compaiono transazione eseguite su entrambi i nodi.
Mi chiedo, nel caso in cui effettivamente il problama sia dovuto a transazioni in conflitto sui due server, se esiste un modo in cui questa tabella può essere "ripulita".

A presto,
Marco

openpj

Per poter far funzionare bene un cluster Alfresco è necessario che tutte le macchine coinvolte a livello di sistema operativo utilizzino lo stesso time server per la sincronizzazione dell'orologia del sistema operativo.
Controlla se effettivamente l'orologio per tutti i nodi è sincronizzato nello stesso modo.