- High latency in metada write
Re: High latency in metada write
Hi, in the test, it writes 1K to DAOS server, the engine actually does:
- Search the dkey in SCM
- Create index for the dkey if it does not exist (b+tree stored in SCM)
- Do the same for akey
- Copy 1K data to SCM
- All the above writes to SCM are in the same PMDK transaction which has its own cost.
This is the reason that VOS write latency is higher than one SCM write. Array write latency should roughly be the same as single value, we will do some benchmark and check if there is any issue.
email@example.com <firstname.lastname@example.org> on behalf of shadow_vector@... <shadow_vector@...>
Date: Thursday, January 27, 2022 at 2:25 PM
To: email@example.com <firstname.lastname@example.org>
Subject: Re: [daos] High latency in metada write
So you mean that although the operation over SCM is fast， but several sequential operations over SCM may result in the high latency？
Here is the test in my server with vos_perf：
Using array type， the latency is much higher(wrose than my test). The type "array" here means to the array in DAOS interface? Is
there something wrong?
Another question, DAOS write the data buffer before the SCM operation in VOS. I think the leader would waite，until follower complete
all the operation and reply and this result in a higher latency. Is there some problem with my understanding?
I'm confuesd with the array update flow now.
Join email@example.com to automatically receive all group messages.