<div dir="ltr"><div style>First off I apologize for the length.</div><div><br></div>There have been a few Jiras recently that have identified L1 consistency issues with both TX and non TX sync caches. Async caches with L1 have their own issues as well, but I only wanted to talk about sync caches.<div>
<br></div><div><a href="https://issues.jboss.org/browse/ISPN-3197">https://issues.jboss.org/browse/ISPN-3197</a><br></div><div><a href="https://issues.jboss.org/browse/ISPN-2965">https://issues.jboss.org/browse/ISPN-2965</a><br>
</div><div><a href="https://issues.jboss.org/browse/ISPN-2990">https://issues.jboss.org/browse/ISPN-2990</a><br></div><div><br></div><div style>I have proposed a solution in <a href="https://github.com/infinispan/infinispan/pull/1922">https://github.com/infinispan/infinispan/pull/1922</a> which should start L1 consistency down the right track. There are quite a few comments on it if you want to look into it more, but because of that I am moving this to the dev mailing list.</div>
<div style><br></div><div style>The key changes in the PR are the following (non-tx):</div><div style><br></div><div style>1. Concurrent reads for a key that can retrieve a remote value are "corralled" into a single thread of execution for that given key. This would reduce network traffic with concurrent gets for the same key. Note the "corralling" only happens on a per key basis.</div>
<div style>2. The single thread that is doing the remote get would update the L1 if able (without locking) and make available the value to all the requests waiting on the get.</div><div style>3. Invalidations that are received would first check to see if there is a current remote get occurring for it's keys. If there is it will attempt to cancel the L1 write(s) before it occurs. If it cannot cancel the L1 write, then it must also wait on the current remote get completion and subsequently run the invalidation. Note the cancellation would fail when the remote get was done and it is in the middle of updating the L1, so this would be very small window.</div>
<div style>4. Local writes will also do the same thing as the invalidation with cancelling or waiting. Note that non tx local writes only do L1 invalidations and don't write the value to the data container. Reasons why I found at <a href="https://issues.jboss.org/browse/ISPN-3214">https://issues.jboss.org/browse/ISPN-3214</a></div>
<div style>5. Writes that require the previous value and don't have it in the L1 would also do it's get operations using the same "corralling" method.<br></div><div style><br></div><div style>4/5 are not currently implemented in PR.</div>
<div style><br></div><div style>This approach would use no locking for non tx caches for all L1 operations. The synchronization point would be done through the "corralling" method and invalidations/writes communicating to it.</div>
<div style><br></div><div style>Transactional caches would do almost the same thing as non-tx. Note these changes are not done in any way yet.</div><div style><br></div><div style>1. Gets would now update the L1 immediately after retrieving the value without locking, but still using the "corralling" technique that non-tx does. Previously the L1 update from a get was transactional. This actually would remedy issue [1]</div>
<div style>2. Writes currently acquire the remote lock when committing, which is why tx caches are able to update the L1 with the value. Writes would do the same cancellation/wait method as non-tx.</div><div style>3. Writes that require the previous value and don't have it in the L1 would also do it's get operations using the same method.</div>
<div style>4. For tx cache [2] would also have to be done.</div><div style><br></div><div style>[1] - <a href="https://issues.jboss.org/browse/ISPN-2965?focusedCommentId=12779780&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-12779780">https://issues.jboss.org/browse/ISPN-2965?focusedCommentId=12779780&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-12779780</a></div>
<div style>[2] - <a href="https://issues.jboss.org/browse/ISPN-1540">https://issues.jboss.org/browse/ISPN-1540</a></div><div style><br></div><div style>Also rehashing is another issue, but we should be able to acquire the state transfer lock before updating the L1 on a get, just like when an entry is committed to the data container.</div>
<div style><br></div><div style>Any comments/concerns would be appreciated.</div><div style><br></div><div style>Thanks,</div><div style><br></div><div style> - Will</div></div>