We currently have ~12 million docs running on our production cluster as follows:
3 total server nodes, c5.4xlarge: (1 data, 1 data/index, 1 data/query)
2 total sync gateways in a load balancer, both c5.4xlarge
When inserting a decent chunk of data via the SG API (~500k docs) the indexing seems to take longer and longer. We recently upped all our instances from m4.xlarges but the re-indexing seems to be snowballing as we put more and more data into the server.
It has been doing this for close to an hour and only has reached 4%. Both of the sync_gateway logs show logs along the line of this:
2018-07-10T12:33:01.413Z Timeout waiting for view “role_access” to be ready for bucket “data” - retrying…
2018-07-10T12:33:01.413Z Timeout waiting for view “access” to be ready for bucket “data” - retrying…
2018-07-10T12:39:03.515Z Timeout waiting for view “channels” to be ready for bucket “data” - retrying…
I’m not quite sure what else we can do at this point to expedite this process. Let me know if you need any other information
Edit:
I went over to the logs in the server UI and I see repetition of some sort of the following:
Compactor for view data/_design/sync_gateway_2.0
(pid [{type,view},
{name,
<<“data/_design/sync_gateway_2.0”>>},
{important,false},
{fa,
{#Fun<compaction_new_daemon.25.86110551>,
[<<“data”>>,
<<"_design/sync_gateway_2.0">>,
{config,
{30,undefined},
{30,undefined},
undefined,false,
false,
{daemon_config,
30,131072,
20971520}},
false,
{[{type,
bucket}]}]}}]) terminated unexpectedly (ignoring this): {not_enough_space,
<<“data”>>,
<<"_design/sync_gateway_2.0">>,
21399369762,
7388863488}