after server reboot: \"...but we are still recovering. Change status ASAP\"

[frankySie]

2 Servers are redundant and one server was rebooted.

The server is running, it has startup completly, but it can't be switched to "active". When it should switch to active then the log shows:
>>> Status change to active requested, but we are still recovering. Change status ASAP.

This status never changes.

Well , at server startup a log-msg is mysterie:
WCCILevent (0), 2018.03.21 14:37:41.419, REDU, SEVERE, 54, Unexpected state, EventManager, evMain, Redundant peer recovery timeout - aborting recovery
WCCILdata (0), 2018.03.21 14:37:41.419, REDU, WARNING, 0, , Recovery request aborted from event.

How I can proceed with analysation?
Thx for response. The complete logs can be sent on request.


Wincc OA 3.12, on Windows Server. 2 Server redundant.

[innovis]

Your mistery message is no mistery at all. For some reason you are spending more time than expected to recover. Check the network connection and the data transfer rate between both servers. Check also the hardware performance of your system. See https://portal.etm.at/index.php?option= ... &Itemid=54 for more information.
If you are still performing bad you can increase your recovery timeout (passiveRecoveryTimeout in your config.redu).
PS: the number of clients doesn't matter at this point, only the size of your database including value archives does.

[frankySie]

Thx very much for your answer. It helped to look at the right configs.

The "passiveRecoveryTimeout" was increased. And it was moved into [data]. Before it was in [event] which is wrong, so the Wincc OA help describes it.
The server was rebooted and it works.

[Gertjan van Schijndel]

Franz Leitner wrote:

Before it was in [event] which is wrong, so the Wincc OA help describes it.

The documentation is correct. The 'passiveRecoveryTimeout' config entry exists in both sections (data and event), but applies to different parts of the recovery.