|
One of our (French) customers is running into this problem under Mobilink 11.0.1.2506. The Mobilink client runs as a service and synchronizes every 1 minute. This works for about 8 to 48 hours, then every cycle fails with "... Internal Error ... " as seen in the ML client log excerpt below. Performing a manual sync succeeds even after the Mobilink client service has failed. Any help understanding the logged error would be appreciated. Mobilink Client Service Configuration (dbmlsync.exe): -c DSN=SM_3;uid=dba;pwd=sql; -n sm3_replication -v -o c:\Replication_Logs\mlclient_app_db_svc_SM_3.txt -os 3m -e sch=EVERY: 0000:01 Log text for one full replication Cycle (logs are in French): I. 2012-06-04 01:06:04. Connexion à la base de données distante I. 2012-06-04 01:06:04. Début de la synchronisation de 'sm3_replication' pour l'utilisateur MobiLink 'ML_CLIENTUSERNAME' I. 2012-06-04 01:06:04. Le balayage du journal démarre à l'offset 072928134613 I. 2012-06-04 01:06:04. En suspension à la fin du journal de transactions actif I. 2012-06-04 01:06:04. Le balayage du journal s'est terminé à l'offset 072928172580 I. 2012-06-04 01:06:04. Vérification de l'activité de la connexion à MobiLink. I. 2012-06-04 01:06:04. Réutilisation de la connexion existante à MobiLink. I. 2012-06-04 01:06:04. Début du téléchargement I. 2012-06-04 01:06:04. Téléchargement des opérations sur la table I. 2012-06-04 01:06:04. Application du téléchargement par MobiLink en attente E. 2012-06-04 01:06:04. Erreur interne : une réponse avec descripteur 44500 a été reçue alors que le descripteur 44495 était attendu (commande = MLCMD_START_ACTIONS). I. 2012-06-04 01:06:04. Ouverture de la connexion à la base de données distante pour appeler un hook d'erreur/de journal. I. 2012-06-04 01:06:04. Fermeture de la connexion pour les hooks d'erreur/de journal. I. 2012-06-04 01:06:04. Fin de la synchronisation de 'sm3_replication' pour l'utilisateur MobiLink 'ML _ CLIENTUSERNAME' I. 2012-06-04 01:06:04. Déconnexion du serveur MobiLink I. 2012-06-04 01:06:05. Balayage complet du journal obligatoire. I. 2012-06-04 01:06:05. Synchronisation terminée I. 2012-06-04 01:06:05. Déconnexion de la base de données distante I. 2012-06-04 01:06:05. Connexion à la base de données distante I. 2012-06-04 01:06:05. Déconnexion de la base de données distante I. 2012-06-04 01:06:05. Prochaine synchronisation planifiée le jui 04 01:07:00am I. 2012-06-04 01:06:05. Connexion à la base de données distante I. 2012-06-04 01:06:05. Vérification du dernier état de synchronisation avec le serveur MobiLink I. 2012-06-04 01:06:05. Connexion au serveur MobiLink sur 'host=172.27.229.68;port=2439;' en utilisant 'TCPIP' I. 2012-06-04 01:06:05. Déconnexion de la base de données distante I. 2012-06-04 01:06:05. Le balayage du journal démarre à l'offset 072928172580 I. 2012-06-04 01:06:05. Connexion à la base de données distante I. 2012-06-04 01:06:05. Traitement des journaux de transactions depuis le répertoire "D:\Database\" I. 2012-06-04 01:06:05. Traitement des transactions depuis le journal de transactions actif I. 2012-06-04 01:06:05. Déconnexion de la base de données distante I. 2012-06-04 01:06:05. En suspension à la fin du journal de transactions actif Thanks, Bob Leviton |
|
This internal error indicates unexpected confusion between the MobiLink client and MobiLink server. This may indicate a bug, or possibly a network intermediary doing something unusual that breaks synchronization. Please make sure the MobiLink client version is less than or equal to the MobiLink server version. If this error persists, you should open a case with technical support. |
|
Another cause could be two (or more) remote databases using the same remote ID, or a remote database being reset or restored from backup to an earlier point in time. Try searching for "mobilink progress offset mismatch" for more info, and if it's still a problem follow Russ' earlier advice to open a case with technical support. Thanks, Graham. As we've seen before, the error stopped after a few days on the customer system. But from experience, it will occur again before long. How likely is it that tech support will be willing to help if we can't provide a reproducible example (we can't reproduce in our lab, even with databases from the customer)?
(29 Jan '13, 13:29)
Bob Leviton
Replies hidden
Tech support is always willing to help. :)
(29 Jan '13, 15:40)
Graham Hurst
|
|
This seems to be solved now. It finally turned out that there was a mismatched version of SQL Anywhere / Mobilink between the consolidated and remote systems. The customer had failed to follow our instructions for bringing every server up to the specified EBF (11.0.1.2506) in the case of the consolidated server. Applying the EBF appears to have corrected the issue. We are now working on procedures to force re-synchronization of much data that was not properly replicated while the problem was occurring. So, in case Russ has given the correct advice
feel free to accept his answer...
(20 Feb '13, 09:29)
Volker Barth
|
This error has returned. It is currently occurring on every sync cycle, causing the cycle to fail. As seen above, the difference between the received and expected descriptor is always 5. Any further insight as to the possible cause would be greatly appreciated.
Thanks, Bob