Not the answer you need?
Register and ask your own question!

Cluster locks, insert/update issues

FaqFaq EntrantCurrent User Role Beginner
Hello,
We've attempted a second migration to Percona cluster and we're having issues realted inserts /updates on nodes.
This is our 3 node cluster configuration: http://wklejto.pl/249433
node1 :
80GB ram
4TB Raid 6 array, hdd 15k [FC 6gbit]
Intel(R) Xeon(R) CPU E5-2420 v2 @ 2.20GHz [24 cores]

node2 :
80GB ram
4TB Raid 6 array, hdd 15k [FC 6gbit]
Intel(R) Xeon(R) CPU E5-2420 v2 @ 2.20GHz [24 cores]

node3 :
80GB ram
4TB Raid 6 array, hdd 15k [FC 6gbit]
Intel(R) Xeon(R) CPU E5-2420 0 @ 1.90GHz [24 cores]


1gbps link between servers


1. making updates on node 2 to table 1, while making insterts on node 1 to table 1 makes database unresponsive
2. making large updates on node 2 in table 2
3. we've also experienced today a total server lock, "BF lock wait long", to unlock we've had to restart first node, and move writing to node 2
Is there anything we can do in this situation when we can't write anything to DB ?

Any ideas what we should change?
Load avarage:
node 2 [writing]: 6
node 1 [some writes, mostly selects]: 2
node 3 [mostly selects]: 2

We have 2 main tables in our DB.
We write to / update T1 on node 2,
and write to / update T2 on node 1

We do this to avoid deadlocks while using same tables on multiple nodes.


Thanks in advance,
Faq

Comments

Sign In or Register to comment.

MySQL, InnoDB, MariaDB and MongoDB are trademarks of their respective owners.
Copyright ©2005 - 2020 Percona LLC. All rights reserved.