Thanks Shawn.
Yes, all Solr nodes know about all three ZK servers (i.e., the zk host string
is of the form zk_a_ip:2181,zk_b_ip:2181,zk_c_ip:2181).
Sorry for the dense description of things: I erred on the side of oversharing
because I didn't want to leave out something useful but I know it makes for an
investment to read so I really appreciate that you took the time. I'm obviously
happy to clarify whatever I can.
I'll make sure to follow the discussion on ZOOKEEPER-2348.
Thanks again,
Marko
If there's anything I can do to make
On 4/26/19, 2:51 PM, "Shawn Heisey" <apa...@elyograg.org> wrote:
On 4/26/2019 9:23 AM, Marko Babic wrote:
> Apologies for bumping my own post but I'm just wondering if it'd be more
appropriate for me to cut a ticket at this point rather than ask the mailing
list.
No, the mailing list is the correct location. If the discussion
determines that there is a bug to be fixed, then it would be appropriate
to open an issue in Jira. For this project, Jira is not used as a
support portal.
Have you informed your Solr server about all three ZK servers in the
zkhost string, or is it only connecting to server A (the one that was
partitioned)?
I'm having a difficult time mentally following the partition description
you sent. Almost certainly a failing on my part, not yours. When I
have a little more free time, I will examine it closer.
From what I've been able to decipher so far, the zookeeper bug that you
linked sounds probable. The bug only mentions 3.5.1 as an affected
version. Solr 7.7.1 includes ZK 3.4.13. I will comment on the bug to
find out whether it affects 3.4 versions as well.
Thanks,
Shawn
