xsengine crashdump and prevents other services starting?

[Former Member]

Long story...

We have been trying to sort out pacemaker clustering and failover, doing a lot of hdbnsutil configuration.

Suddenly we are unable to start either primary or secondary (even after topology cleanup).

The services appear to try to start, however index server gets hung up trying to contact xsengine on 30207.  xsengine constantly crashdumps.

I can't get into studio in full admin mode to look at global.ini, but we've done nothing at all to xsserver.  The only change made today was log_backuptimeout=0 per a document from Cisco.

This does exist in the custom global.ini, but since the systems don't like to start or stop cleanly (they hang endlessly with a started nameserver, daemon.

Index server trace ends with repetition of this:

ERROR: Failed to open channel 127.0.0.1:30207! reason: (connection refused).

xsengine generates a crashdump every time we try to restart.  Whatever the problem is has been replicated to the secondary.  Both systems exhibit the exact same behavior.

is xsengine required to start the index server? If not, is there any way I can disable it and attempt to recover it from a functioning system?

xsengine_alert_ps2han08.trc.zip

[former_member183326]

Hi James,

I would re-initialize the Xsengine? Do you use the Xsengine?

It also is matching a bug with a permanent solution being to update.

2306382 - Crash On Secondary After Takeover With The Error: "log position inconsistency detected - p...

Is re-initializing an option for you? You could try removing the XS from the landscape and adding it again:

1697613 - Removing SAP HANA XS from topology

[former_member182967]

Hi James,

I see the following error in the trace file you provided:

exception  1: no.2100002  (ptime/storage/recovery/RowStoreTransactionCallback.cc:446)

    log position inconsistency detected - please contact HANA dev support before clearing or throwing away your redo log files

Please check if the notes below are helpful or not.

2306382 - Crash On Secondary After Takeover With The Error: "log position inconsistency detected - plz contact to HANA dev support before clearing or throwing away your redo log files"

2129651 - Indexserver crash caused by inconsistent log position when startup

2184052 - SAP HANA: Secondary Side Crashes After Takeover due to "log position inconsistency detected"

Regards,

Ning