GitLab

Essentially reverts part of cd8ff376 to
make sure there is always a contextCSN if the server is in charge of its
own serverID.

Howard Chu authored Feb 22, 2018 and Quanah Gibson-Mount committed Feb 28, 2018

And try another approach. Always write contextCSN updates, but
don't set dont_replicate for updates we want propagated.

Howard Chu authored Feb 02, 2018 and Quanah Gibson-Mount committed Feb 09, 2018

Let the entryCSN check do all the work. Reloading a server from an old
backup needs this to go thru.

Howard Chu authored Jan 30, 2018 and Quanah Gibson-Mount committed Feb 09, 2018

Update syncprov contextCSNs when context entry is added.
Fix accesslog to properly tag Add op when adding context entry.

Ondřej Kuzník authored Aug 25, 2017 and Quanah Gibson-Mount committed Sep 12, 2017

When a checkpoint happens, if we remove the CSN from the pending list,
accesslog won't pass it onto the accesslog DB. But in a delta-mmr
scenario, an accesslog entry without a CSN faces a race where it might
be applied twice - that usually fails and causes a full refresh, other
times it can cause a silent desync - both are undesirable.

Howard Chu authored Feb 02, 2016 and Quanah Gibson-Mount committed Feb 02, 2016

Must setup psearch before snapshotting ctxcsn

Howard Chu authored Jan 22, 2016 and Quanah Gibson-Mount committed Jan 28, 2016

Delay mutex init to avoid leaking the mutex

Howard Chu authored Jan 22, 2016 and Quanah Gibson-Mount committed Jan 28, 2016

Check for abandon just before recording psearch

Howard Chu authored Oct 23, 2015 and Quanah Gibson-Mount committed Oct 26, 2015

Prevent spurious contextCSN generation
and ignore consumers when we have no contextCSN yet.
But make sure to propagate valid contextCSN updates to
accesslog/syncprov for delta consumers.

Howard Chu authored Mar 21, 2015 and Quanah Gibson-Mount committed Mar 23, 2015

Prev patch broke underlying assumption that mods queue and execute
in order. Now must search list for matching mod to dequeue.

Howard Chu authored Jan 31, 2015 and Quanah Gibson-Mount committed Jan 30, 2015

Don't overload ri_csn with cookie, use a separate berval

Howard Chu authored Jan 30, 2015 and Quanah Gibson-Mount committed Jan 30, 2015

Rewrote the psearch result handling to keep track of which
queues are using which results.

Leonid Yuriev authored Jan 04, 2015 and Quanah Gibson-Mount committed Jan 05, 2015

Detaching a syncops record from op-list conditionally,
only when it was freed by syncprov_free_syncop.

The syncprov_free_syncop() and syncprov_drop_psearch() now
returns a flag, which is nonzero if the given syncops was freed.