Updating oplog, reverse order?

[Chris]

Hi,

Due to an issue with a faulty mapreduce job breaking our replication
(error "13629 can't have undefined in a query expression"), we have
been manually updating the oplog to set the broken operations to
"op" : "n". However, with > 120m objects in the oplog, this obviously
takes a long time.

Is there any way to speed up the operation? The faulty operations all
took place within the last 24hrs, so obviously if there was a way to
reverse the order of the scan it would take a lot less time.

Many thanks for your help.

[Sam Millman]

Would it not just be more prudent to fix the mr?

--
You received this message because you are subscribed to the Google Groups "mongodb-user" group.
To post to this group, send email to mongod...@googlegroups.com.
To unsubscribe from this group, send email to mongodb-user...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/mongodb-user?hl=en.

[Chris]

Hi Sam,

Sorry I should have it more clear; the mr issue has been fixed, so no
additional "broken" operations are going into the oplog. However, the
existing entries caused the replication to halt at that point, and
until they're removed / set as op : n, the replication can't continue.
So, as I've done once before, I'm manually updating these entries so
that replication can resume.

Cheers

On Nov 3, 12:59 pm, Sam Millman <sam.mill...@gmail.com> wrote:
> Would it not just be more prudent to fix the mr?
>