<div dir="ltr"><div dir="ltr">Hi <a class="gmail_plusreply" id="gmail-plusReplyChip-0" href="mailto:stephen@networkplumber.org" tabindex="-1">@Stephen Hemminger</a> ,<div><br></div><div>Thanks for the reply , please find the details below.</div><div><br></div><div>DPDK version : 20.11.6</div><div><br>#define LOADBAL_HASH_ENTRIES_MAX (1024*1024*36)<br>#define LOADBAL_HASH_TABLE_SIZE_MULTIPLIER 2<br><br>hashSizeMultiplier = LOADBAL_HASH_TABLE_SIZE_MULTIPLIER;<br><br> struct rte_hash_parameters loadbal_hash_params = {<br> .name = NULL,<br> .entries = LOADBAL_HASH_ENTRIES_MAX * hashSizeMultiplier,<br> .key_len = sizeof(flow_key_t),<br> .hash_func = app_hash_crc(internally it calls rte_hash_crc_4byte for v4/v6)<br> .hash_func_init_val = 0,<br> };<br><br>First, what is the return value, which error?<br> we captured only the return value, and will check what error it is returning.<br><br>IPv4 Load-Bal Flow hash table:<br> numInsertions: 998214247<br> numInsertionsFailures: 4252<br> numRemovals: 997197485<br> numRemovalFailures: 214902<br> numObjects: 1016762<br> NumBuckets: 75497472<br> TableCapacity: 75497472<br> LoadFactor: 1%<br> NumLookupSuccess: 389165605<br> NumLookupFails: 2704814006<br> Failure Analysis:<br> TableFull(>=95%): 0<br> HighLoad(75-95%): 0<br> MediumLoad(50-75%): 0<br> LowLoad(<50%): 4252<br> </div></div>We are try to analyze the code and find out the details,<div>Initially there will be no issues, insertion starts failing after 24 HRS or so.</div><div><br></div><div>Thanks,</div><div>Venkatesh.</div><div><br><div><br><div class="gmail_quote gmail_quote_container"><div dir="ltr" class="gmail_attr">On Sun, Jun 8, 2025 at 9:17 PM Stephen Hemminger <<a href="mailto:stephen@networkplumber.org">stephen@networkplumber.org</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">On Sun, 8 Jun 2025 00:36:02 +0530<br>
venkatesh bs <<a href="mailto:venki.bsv@gmail.com" target="_blank">venki.bsv@gmail.com</a>> wrote:<br>
<br>
> Hi @dpdk community,<br>
> <br>
> In my application, I am using the dpdk hash table as below.<br>
> <br>
> 1. calculate the hash value.<br>
> 2. using this signature add the key and data into a table.<br>
> 3. lock is used in the 2nd call (add).<br>
> <br>
> Below is the code sniffer for the same.<br>
> ==================================================================<br>
> hash_sig_t sig = rte_hash_hash(hash_tablele, (void *) &new_key);<br>
> <br>
> pthread_mutex_lock(&lock);<br>
> int32_t ret = rte_hash_add_key_with_hash_data(hash_table,<br>
> (void *)&new_key,<br>
> sig,<br>
> info);<br>
> <br>
> if (ret < 0)<br>
> {<br>
> pthread_mutex_unlock(lock);<br>
> return TOS_E_FAIL;<br>
> }<br>
> pthread_mutex_unlock(lock);<br>
> <br>
> return OK..<br>
> ==================================================================<br>
> <br>
> My application is having a lot of threads and when run with heavy load , I<br>
> am getting a lot of insertion failure, i felt the reason could be<br>
> calculating and adding the has value in 2 api's as opposed to<br>
> rte_hash_add() that is safe under a lock..<br>
> <br>
> <br>
> Please let me know your thoughts.<br>
<br>
Which architecture and DPDK version?<br>
What flags did you use during hash creation?<br>
<br>
As always with open source, the first thing to do is look at the source<br>
and see what is really happening, rather than just relying on the documentation.<br>
<br>
I assume you are using the default hash function which is CRC.<br>
<br>
First what is the return value, which error?<br>
It might just be key collisions. How big is the table and how many inserts?<br>
<br>
Also, the current DPDK hash flags with better locking and RCU.<br>
This would be faster than simple pthread mutex.<br>
<br>
</blockquote></div></div></div></div>