Re: [Samba] [ctdb]Unable to run startrecovery event
- Date: Thu, 6 Sep 2018 18:58:34 +1000
- From: Martin Schwenke via samba <samba@xxxxxxxxxxxxxxx>
- Subject: Re: [Samba] [ctdb]Unable to run startrecovery event
On Thu, 6 Sep 2018 14:24:12 +0800 (CST), "zhu.shangzhong--- via samba"
> I have checked more logs.
> Before ctdb-eventd went away, system memory utilization is very high, almost 100%.
> Is it related to "Bad talloc magic value - wrong talloc version used/mixed"?
Definitely. The only time we have seen this in eventd is when a memory
allocation failed. The fix is commit
e4a5d610b8e81c78b7d98217bc87c4b815b4c4e7. However, I don't think we
backported this because in the same situation we saw several other
out-of-memory symptoms that are more difficult to fix, so it seemed
This fix did make it into 4.9, which should be released next week.
If you really would like to see a backport to other releases then
please open a bug at https://bugzilla.samba.org/ . However, please note
that 4.6 is now security fixes only and this isn't security related.
When 4.9 is released then 4.7 will also be security fixes only.
I'm still working on the recovery lock issue...
peace & happiness,
To unsubscribe from this list go to the following URL and read the