I am in agreement with Patrick, this is a typical symptom of saturated IO. Are there a high of drops and/or pending compactions?
Get Outlook for Android<https://aka.ms/ghei36> ________________________________ From: Patrick McFadin <pmcfa...@gmail.com> Sent: Tuesday, January 28, 2020 11:25:49 AM To: user@cassandra.apache.org <user@cassandra.apache.org> Subject: Re: How to read content of hints file and apply them manually? Just to add in here. Any time I see any hints on a cluster, that's like seeing smoke. If you can't explain it, you have a fire somewhere and it's not going to get any better. By the few messages I've seen, I would start by looking at your IO subsystem on your nodes. Do you have enough throughput to write and read at the same time? These are exactly the symptoms I see when running Cassandra on a SAN or NAS. Patrick On Mon, Jan 27, 2020 at 8:17 PM Surbhi Gupta <surbhi.gupt...@gmail.com<mailto:surbhi.gupt...@gmail.com>> wrote: We tried to tune sethintedhandoffthrottlekb to 100 , 1024 , 10240 but nothing helped . Our hints related parameters are as below, if you don't find any parameter below then it is not set in our environment and should be of the default value. max_hint_window_in_ms: 10800000 # 3 hours hinted_handoff_enabled: true hinted_handoff_throttle_in_kb: 100 max_hints_delivery_threads: 8 hints_directory: /var/lib/cassandra/hints hints_flush_period_in_ms: 10000 max_hints_file_size_in_mb: 128 On Mon, 27 Jan 2020 at 18:34, Jeff Jirsa <jji...@gmail.com<mailto:jji...@gmail.com>> wrote: The high cpu is probably the hints getting replayed slamming the write path Slowing it down with the hint throttle may help It’s not instant. On Jan 27, 2020, at 6:05 PM, Erick Ramirez <flightc...@gmail.com<mailto:flightc...@gmail.com>> wrote: Increase the max_hint_window_in_ms setting in cassandra.yaml to more than 3 hours, perhaps 6 hours. If the issue still persists networking may need to be tested for bandwidth issues. Just a note of warning about bumping up the hint window without understanding the pros and cons. Be aware that doubling it means: * you'll end up doubling the size of stored hints in the hints_directory * there'll be twice as much hints to replay when node(s) come back online There's always 2 sides to fiddling with the knobs in C*. Cheers!
