D
D
D_E_S2016-05-13 17:28:45
PostgreSQL
D_E_S, 2016-05-13 17:28:45

What to do if postgresql loads hdd at 100%?

There is only one site per server. Recently, the system began to fail. The number of users has not changed, the requests have not changed.
iostat
Device: rrqm/s wrqm/sr/sw/s rkB/s wkB/s avgrq-sz avgqu-sz await r_await w_await svctm %util
xvda 0.00 0.00 760.80 1.30 16034.80 57.20 42.23 165.32 68.49 37.18 18393.54 1.31 100.00
if you assure the postgresql service, then the load on the hdd:
Device: rrqm/s wrqm/sr/sw/s rkB/s wkB/s avgrq- sz avgqu-sz await r_await w_await svctm %util
xvda 0.00 4.80 0.20 2.20 1.20 39.60 34.00 0.00 0.00 0.00 0.00 0.00 0.00
Logging is temporarily disabled. Decreasing or increasing shared_buffers did not help

Answer the question

In order to leave comments, you need to log in

1 answer(s)
L
lega, 2016-05-13
@D_E_S

postgresql is doing a big rkB/s, probably there is a data fullscan, i.e. indexes are not used - look for requests to add the correct indexes.

Didn't find what you were looking for?

Ask your question

Ask a Question

731 491 924 answers to any question