Write counter (“wr”) is not going down


#1

Hi, We have a new cluster installed in AWS and are seeing some issues.

$ asmonitor -e "latency -k reads"

3 hosts in cluster: ***
 
        ====reads====
                          timespan   ops/sec    >1ms    >8ms   >64ms
***:3000    20:22:38-GMT->20:22:48     217.4   90.85   88.22   66.24
***:3000    20:22:45-GMT->20:22:55     192.8    2.96    0.00    0.00
***:3000    20:22:45-GMT->20:22:55     221.5   67.36   62.75   16.79

I’ve been troubleshooting this for a while and noticed that on this page it states:

If the write counter (“wr”) is not going down, it means requests are getting backed up. check for growth pattern

So, I checked and this is what I’m seeing on one of the “bad” nodes:

$ sudo tail -200f /var/log/aerospike/aerospike.log |grep trans_in_progress
Oct 07 2014 20:17:38 GMT: INFO (info): (thr_info.c::4416)    trans_in_progress: wr 14 prox 0 wait 0 ::: q 17 ::: bq 0 ::: iq 0 ::: dq 0 : fds - proto (52, 8894, 8842) : hb 3 : fab 44
Oct 07 2014 20:17:48 GMT: INFO (info): (thr_info.c::4416)    trans_in_progress: wr 16 prox 0 wait 0 ::: q 34 ::: bq 0 ::: iq 0 ::: dq 0 : fds - proto (52, 8894, 8842) : hb 3 : fab 44
Oct 07 2014 20:17:58 GMT: INFO (info): (thr_info.c::4416)    trans_in_progress: wr 14 prox 0 wait 0 ::: q 27 ::: bq 0 ::: iq 0 ::: dq 0 : fds - proto (52, 8894, 8842) : hb 3 : fab 44
Oct 07 2014 20:18:08 GMT: INFO (info): (thr_info.c::4416)    trans_in_progress: wr 15 prox 0 wait 0 ::: q 31 ::: bq 0 ::: iq 0 ::: dq 0 : fds - proto (52, 8898, 8846) : hb 3 : fab 44

And I’m seeing this on the “good” nodes:

$ sudo tail -200f /var/log/aerospike/aerospike.log |grep trans_in_progress
Oct 07 2014 19:59:45 GMT: INFO (info): (thr_info.c::4416)    trans_in_progress: wr 0 prox 0 wait 0 ::: q 0 ::: bq 0 ::: iq 0 ::: dq 0 : fds - proto (15, 8480, 8465) : hb 3 : fab 44
Oct 07 2014 19:59:55 GMT: INFO (info): (thr_info.c::4416)    trans_in_progress: wr 0 prox 0 wait 0 ::: q 0 ::: bq 0 ::: iq 0 ::: dq 0 : fds - proto (15, 8480, 8465) : hb 3 : fab 44
Oct 07 2014 20:00:05 GMT: INFO (info): (thr_info.c::4416)    trans_in_progress: wr 0 prox 0 wait 0 ::: q 0 ::: bq 0 ::: iq 0 ::: dq 0 : fds - proto (15, 8480, 8465) : hb 3 : fab 44
Oct 07 2014 20:00:15 GMT: INFO (info): (thr_info.c::4416)    trans_in_progress: wr 0 prox 0 wait 0 ::: q 0 ::: bq 0 ::: iq 0 ::: dq 0 : fds - proto (15, 8480, 8465) : hb 3 : fab 44

What’s interesting is that I’m not even doing any writes or deletes. I’m only doing reads. I’m not sure where to go from here with the troubleshooting. Any help would be appreciated.

Thanks, Ben


#2

Hi Ben,

I’ve responded back to you on your ticket that you have submitted to us. Also please remember since you are a Enterprise support, please feel free to submit your ticket through there at any time.

But forum is another great way to ask for assistance or guidance from the community members.

Jerry


#3

Thanks for the quick response, Jerry. I’ll continue the conversation via email.