Hi again, Today I have started the same set of tests for cb server; currently I have been adding data to server with a small rate (200/sec), my stats look like the below
Every 2.0s: /opt/couchbase/bin/cbstats localhost:11210 -b default all | > egrep 'item|mem|flusher|ep_queue|bg|eje|resi|warm' Tue Jan 7 > 08:45:05 2014 curr_items: 40798244 > curr_items_tot: 49348787 > curr_temp_items: 0 > ep_access_scanner_num_items: 0 > ep_bg_fetch_delay: 0 > ep_bg_fetched: 0 > ep_bg_meta_fetched: 0 > ep_bg_remaining_jobs: 0 > ep_chk_max_items: 5000 > ep_diskqueue_items: 1812316 > ep_diskqueue_memory: 57994112 > ep_failpartialwarmup: 0 > ep_flusher_state: running > ep_flusher_todo: 115917 > ep_item_begin_failed: 0 > ep_item_commit_failed: 0 > ep_item_flush_expired: 0 > ep_item_flush_failed: 0 > ep_item_num_based_new_chk: 1 > ep_items_rm_from_checkpoints: 66 > ep_max_item_size: 20971520 > ep_mem_high_wat: 12777527705 > ep_mem_low_wat: 11274289152 > ep_mem_tracker_enabled: true > ep_meta_data_memory: 4920394995 > ep_mutation_mem_threshold: 95 > ep_num_eject_failures: 1111132682 > ep_num_non_resident: 47445563 > ep_num_value_ejects: 10029584 > ep_queue_size: 1812316 > ep_tap_backfill_resident: 0.9 > ep_tap_bg_fetch_requeued: 0 > ep_tap_bg_fetched: 6120426 > ep_tap_bg_max_pending: 500 > ep_total_del_items: 0 > ep_total_new_items: 6759834 > ep_uncommitted_items: 115917 > ep_waitforwarmup: 0 > ep_warmup: 1 > ep_warmup_batch_size: 1000 > ep_warmup_dups: 0 > ep_warmup_min_items_threshold: 100 > ep_warmup_min_memory_threshold: 100 > ep_warmup_oom: 0 > ep_warmup_thread: complete Do you note anything unusual? Some of the inserts are backing off with retries from my client implementation. Some of my objects require to retry for about 20-30 times till they are accepted by the server. From the stats i find that the queue is a bit large with this rate I insert data.. And i cannot understand the ep_num_eject_failures which is constantly increasing, is it some sort of a counter? for insert failures or something similar? Thanks On Friday, 3 January 2014 14:52:50 UTC+2, Thomas wrote: > > Hi, > > I have been performing some tests to couchbase server 2.1.1 community > edition and it seems that I have an issue with data insertion. I have been > following the instructions from the following URL: > http://docs.couchbase.com/couchbase-sdk-java-1.2/#advanced-usage > > The problem is that it seems that for a very large period in time (e.g. > more than two hours) the couchbase freezes and does not accept new > documents via the Java client (set(...)) making it problematic in data > insertion > > Any ideas of why this is happening? > > My cluster is three nodes of 42GB memory in total and I have currently > about 100million documents in one bucket (default) > > Let me know if you need any additional information, from my cluster > > Thanks > -- You received this message because you are subscribed to the Google Groups "Couchbase" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/groups/opt_out.
