Why jit can destroy performance on Linux server? - performance

We have a Postgresql development server on Windows and a production server on Ubuntu Linux. The Postgresql release versions are slightly different, the newer in on linux (Postgresql 14.5).
One database was exported from development and imported in production, so they are the same. We noticed that some queries are 100 times slower on Ubuntu server. Too much to blame a different hardware. The critical queries are like the following:
SELECT a.field1,
...
a.field_n,
CASE (EXISTS (
SELECT b."waterBodyCode"
FROM "GAP_ecologico_SWB_1" b
WHERE a.field1 = b.field1_b AND ...))
WHEN true
THEN 'Yes'
ELSE 'No'
END AS "attributableQE"
FROM "SWB_ecological_quality_elements" a
ORDER BY a.field1, ...;
I know we can rewrite the query in a better and more efficient way but before doing that we would like to understand the root cause of the poor performance on Ubuntu.
SWB_ecological_quality_elements and swaterBodyCode are views. What we suspect is that on Windows server the inner query is executed using indexing or caching while on Ubuntu a full scan is done. The postgresql.conf files are quite similar, what can we check ?
UPDATE 1
I tuned the memory as suggested by PgTune but it was useless, same poor performance. I executed the query with Analyze, the timing result looks weird but I'm not an expert.
" Functions: 1499"
" Options: Inlining true, Optimization true, Expressions true, Deforming true"
" Timing: Generation 101.424 ms, Inlining 103.476 ms, Optimization 10402.780 ms, Emission 6866.288 ms, Total 17473.968 ms"
"Execution Time: 18005.457 ms"
UPDATE 2
Setting jit to false cause an incredible performance boost on the query execute on the Linux server. Don't know anything about jit and cannot figure out why it behaves so differently on Windows (Postgresql 14.4) and on Linux (Postgresql 14.5), but others noticed its weird effects as can be read here: Handling performance problems with jit in postgres 12.
Any explanation ?

Related

VB6 compiled is slow when copying files

I know, VB6 is historic...ok, but...
I w
rote years ago a backup program not being satisfied from coomercial producuts I tested.
Now I wanted to renew it with some enhancements and a new graphic; the result is quite good for me. Since the file copying process is generally rather slow, I thought to compile it to squeeze some seconds...and instead...this is much slower.
Here are some info:
Win10-64 (version 22H2 just upgraded)
Tested on the same PC with identical parameters
VB6 runs with admin privileges, in Win7 SP3 compatibility mode.
Even if it is not relevant here, the job was to copy a folder containing other 426 folders and 4598 files of different sizes (from 1kB to 435MB, for a total of 1.05GB), from an inside SSD disk to an external SSD disk.
The interpreted version took 7.2 sec while the compiled version ended in 18.6 sec !
I tried different compilation setting in native code, dismissing all the advanced controls over ranges, integers and floats, without any notable difference.
I could accept a small difference for some unknown reason, but it is unreal to get a 2.5:1 ratio.
Any idea?
EDIT
Based on comments:
I repeated the comparison several times; the differences (in both the compiled and the interpreted mode) is around +/- 1sec.
Files are copied using filesystemobject.copyfile
my admin privileges are the same for both
Again, I'm not complaining nor worried by the absolute time the copy takes, I can survive with that since it is an operation made every week and during easy hours.
What is surprising is WHY it happens.
Even the idea to compile the program was due to my curiosity since there is very little to optimize in the code; it is just a for-next loop with very little calculations and assignements.
The program takes the dir and files info from a text-based DB created by recursively scanning of the source folder, then loaded into a custom array...pretty simple.
This is done before the actual copy phase, which is what I'm investigating.

Neo4j in Docker - Max Heap Size Causes Hard crash 137

I'm trying to spin up a Neo4j 3.1 instance in a Docker container (through Docker-Compose), running on OSX (El Capitan). All is well, unless I try to increase the max-heap space available to Neo above the default of 512MB.
According to the docs, this can be achieved by adding the environment variable NEO4J_dbms_memory_heap_maxSize, which then causes the server wrapper script to update the neo4j.conf file accordingly. I've checked and it is being updated as one would expect.
The problem is, when I run docker-compose up to spin up the container, the Neo4j instance crashes out with a 137 status code. A little research tells me this is a linux hard-crash, based on heap-size maximum limits.
$ docker-compose up
Starting elasticsearch
Recreating neo4j31
Attaching to elasticsearch, neo4j31
neo4j31 | Starting Neo4j.
neo4j31 exited with code 137
My questions:
Is this due to a Docker or an OSX limitation?
Is there a way I can modify these limits? If I drop the requested limit to 1GB, it will spin up, but still crashes once I run my heavy query (which is what caused the need for increased Heap space anyway).
The query that I'm running is a large-scale update across a lot of nodes (>150k) containing full-text attributes, so that they can be syncronised to ElasticSearch using the plug-in. Is there a way I can get Neo to step through doing, say, 500 nodes at a time, using only cypher (I'd rather avoid writing a script if I can, feels a little dirty for this).
My docker-compose.yml is as follows:
---
version: '2'
services:
# ---<SNIP>
neo4j:
image: neo4j:3.1
container_name: neo4j31
volumes:
- ./docker/neo4j/conf:/var/lib/neo4j/conf
- ./docker/neo4j/mnt:/var/lib/neo4j/import
- ./docker/neo4j/plugins:/plugins
- ./docker/neo4j/data:/data
- ./docker/neo4j/logs:/var/lib/neo4j/logs
ports:
- "7474:7474"
- "7687:7687"
environment:
- NEO4J_dbms_memory_heap_maxSize=4G
# ---<SNIP>
Is this due to a Docker or an OSX limitation?
NO Increase the amount of available RAM to Docker to resolve this issue.
Is there a way I can modify these limits? If I drop the requested
limit to 1GB, it will spin up, but still crashes once I run my heavy
query (which is what caused the need for increased Heap space
anyway).
The query that I'm running is a large-scale update across a lot of
nodes (>150k) containing full-text attributes, so that they can be
syncronised to ElasticSearch using the plug-in. Is there a way I can
get Neo to step through doing, say, 500 nodes at a time, using only
cypher (I'd rather avoid writing a script if I can, feels a little
dirty for this).
N/A This is a NEO4J specific question. It might be better to seperate this from the Docker questions listed above.
3.The query that I'm running is a large-scale update across a lot of nodes (>150k) containing full-text attributes, so that they can be syncronised to ElasticSearch using the plug-in. Is there a way I can get Neo to step through doing, say, 500 nodes at a time, using only cypher (I'd rather avoid writing a script if I can, feels a little dirty for this).
You can do this with the help of apoc plugin for neo4j, more specifically apoc.periodic.iterate
or apoc.periodic.commit
.
If you will use apoc.periodic.commit your first match should be specific like in example you mark which nodes have you already synced, because it sometimes fall in the loop:
call apoc.periodic.commit("
match (user:User) WHERE user.synced = false
with user limit {limit}
MERGE (city:City {name:user.city})
MERGE (user)-[:LIVES_IN]->(city)
SET user.synced =true
RETURN count(*)
",{limit:10000})
If you use apoc.periodic.iterate you can run it in parallel mode:
CALL apoc.periodic.iterate(
"MATCH (o:Order) WHERE o.date > '2016-10-13' RETURN o",
"with {o} as o MATCH (o)-[:HAS_ITEM]->(i) WITH o, sum(i.value) as value
CALL apoc.es.post(host-or-port,index-or-null,type-or-null,
query-or-null,payload-or-null) yield value return *", {batchSize:100, parallel:true})
Note that there is no need for second MATCH clause and apoc.es.post is a function for apoc that can send post requests to elastic search.
see documentation for more info

Rcpp in Rstudio, can't cache in memory when parallel if I don't open the cpp file in Rstudio

I met a wired problem but I wonder if I'm asking the correct question:
result = parLapply(cl, 1:4,
function(j,rho_list_needed,delta0_needed,
V_iter_s,Sigma_list_needed) {
rhoj = rho_list_needed[[j]]
delta0_in_cpp = delta0_needed
v = as.vector(V_iter_s[,,,j])
sigmaj = Sigma_list_needed[[j]]
sourceCpp('sample_Z.cpp')#first time complie slow,then cashed
return(Sample_Z(rhoj,delta0_in_cpp, v,sigmaj,A,Cmatrix))
},rho_list_needed,delta0_needed,
V_iter[[s]],Sigma_list_needed)
When I was testing my sample_Z.cpp with parallel through parLapply, the single calculation takes around 1 sec. By parallel, my 4 iterations takes around 1.2 secs, which is a big improvement compared to unparalleled version, which is 8 sec.
There's no problem at all when I run my program yesterday. Just now I noticed a bug and revised my program. To give my PC a fresh environment, I restarted my computer. When started to run my program, I only opened the .R file, and run. But it took 9 sec for that parallel, which used to be 1.2 sec. The 9 sec was after warming up my cores, i.e., already sourced the cpp before I time it.
I just don't know where is the bug. Then try to source the cpp file directly in my global merriment, and I found out that there was no caching at all. The second time took the same time as the first one.
But I accidentally opened the sample_Z.cpp in Rstudio, explicitly at the editor. And then, everything works correct now.
I don't know how to search this similar problem on google with what kind of key words and I don't know if opening the cpp file is a must, while I never known before.
Can anyone tell me what's the real issue? Thanks!
After restarting your PC, you probably had extra processes running which would have competed for CPU cores that slowed down your algorithm. The fact you're rebooting suggests to me you're not using Linux... but if you are, watch with top while starting your code, or equivalent for your platform.

What is an efficient way to access large-ish datasets using JDBC with Clojure

EDIT
N00b problem as it turns out. I didn't realize that running lein ring server results in your app being ran in interpreted mode, which is why it was much slower.
Can the following Clojure/ JDBC fragment be optimized so that it runs (much) faster?
(defn test-sql []
(sql/with-connection (db-connection)
(sql/with-query-results results ["select * from users order by username asc"]
(doseq [row results ]
(println "User" (row :first_name) (row :last_name)) results))))
I am considering using Clojure for an ETL project. First test I wrote was to print out data from a table I have with ~280K records in it. The implementations I came up with so far have been quite slow; what takes ~12 seconds in Java (even using myBatis to populate objects rather than 'raw' access) takes ~9.5 minutes with my Clojure solution.
I tried map instead of doseq, and tried using a cursor like outlined here: http://asymmetrical-view.com/2010/10/14/clojure-and-large-result-sets.html, but I get about the same execution time for each.
FWIW, same result when doing .println java.lang.System/out (not surprising), and using with-query-results*:
(defn test-sql2 []
(sql/with-connection (db-connection)
(sql/with-query-results* ["select * from users order by username asc"]
(fn [row] (println "User" (row :first_name) (row :last_name))))))
same, same.
Well this is embarrassing... As it turns out, the relatively poor performance was due to running my test code as part of a simple web app (I know, that doesn't make sense in the context of this question), which I ran with lein ring server which I guess is the same as running it using repl (I just didn't make that connection). When I tried compiling and packaging with lein uberjar and then executed that jar with java -jar, it gave me comparable performance to the Java app.

I/O performance of multiple JVM (Windows 7 affected, Linux works)

I have a program that creates a file of about 50MB size. During the process the program frequently rewrites sections of the file and forces the changes to disk (in the order of 100 times). It uses a FileChannel and direct ByteBuffers via fc.read(...), fc.write(...) and fc.force(...).
New text:
I have a better view on the problem now.
The problem appears to be that I use three different JVMs to modify a file (one creates it, two others (launched from the first) write to it). Every JVM closes the file properly before the next JVM is started.
The problem is that the cost of fc.write() to that file occasionally goes through the roof for the third JVM (in the order of 100 times the normal cost). That is, all write operations are equally slow, it is not just one that hang very long.
Interestingly, one way to help this is to insert delays (2 seconds) between the launching of JVMs. Without delay, writing is always slow, with delay, the writing is slow aboutr every second time or so.
I also found this Stackoverflow: How to unmap a file from memory mapped using FileChannel in java? which describes a problem for mapped files, which I'm not using.
What I suspect might be going on:
Java does not completely release the file handle when I call close(). When the next JVM is started, Java (or Windows) recognizes concurrent access to that file and installes some expensive concurrency handler for that file, which makes writing expensive.
Would that make sense?
The problem occurs on Windows 7 (Java 6 and 7, tested on two machines), but not under Linux (SuSE 11.3 64).
Old text:
The problem:
Starting the program from as a JUnit test harness from eclipse or from console works fine, it takes around 3 seconds.
Starting the program through an ant task (or through JUnit by kicking of a separate JVM using a ProcessBuilder) slows the program down to 70-80 seconds for the same task (factor 20-30).
Using -Xprof reveals that the usage of 'force0' and 'pwrite' goes through the roof from 34.1% (76+20 tics) to 97.3% (3587+2913+751 tics):
Fast run:
27.0% 0 + 76 sun.nio.ch.FileChannelImpl.force0
7.1% 0 + 20 sun.nio.ch.FileDispatcher.pwrite0
[..]
Slow run:
Interpreted + native Method
48.1% 0 + 3587 sun.nio.ch.FileDispatcher.pwrite0
39.1% 0 + 2913 sun.nio.ch.FileChannelImpl.force0
[..]
Stub + native Method
10.1% 0 + 751 sun.nio.ch.FileDispatcher.pwrite0
[..]
GC and compilation are negligible.
More facts:
No other methods show a significant change in the -Xprof output.
It's either fast or very slow, never something in-between.
Memory is not a problem, all test machines have at least 8GB, the process uses <200MB
rebooting the machine does not help
switching of virus-scanners and similar stuff has no affect
When the process is slow, there is virtually no CPU usage
It is never slow when running it from a normal JVM
It is pretty consistently slow when running it in a JVM that was started from the first JVM (via ProcessBuilder or as ant-task)
All JVMs are exactly the same. I output System.getProperty("java.home") and the JVM options via RuntimeMXBean RuntimemxBean = ManagementFactory.getRuntimeMXBean(); List arguments = RuntimemxBean.getInputArguments();
I tested it on two machines with Windows7 64bit, Java 7u2, Java 6u26 and JRockit, the hardware of the machines differs, though, but the results are very similar.
I tested it also from outside Eclipse (command-line ant) but no difference there.
The whole program is written by myself, all it does is reading and writing to/from this file, no other libraries are used, especially no native libraries. -
And some scary facts that I just refuse to believe to make any sense:
Removing all class files and rebuilding the project sometimes (rarely) helps. The program (nested version) runs fast one or two times before becoming extremely slow again.
Installing a new JVM always helps (every single time!) such that the (nested) program runs fast at least once! Installing a JDK counts as two because both the JDK-jre and the JRE-jre work fine at least once. Overinstalling a JVM does not help. Neither does rebooting. I haven't tried deleting/rebooting/reinstalling yet ...
These are the only two ways I ever managed to get fast program runtimes for the nested program.
Questions:
What may cause this performance drop for nested JVMs?
What exactly do these methods do (pwrite0/force0)? -
Are you using local disks for all testing (as opposed to any network share) ?
Can you setup Windows with a ram drive to store the data ? When a JVM terminates, by default its file handles will have been closed but what you might be seeing is the flushing of the data to the disk. When you overwrite lots of data the previous version of data is discarded and may not cause disk IO. The act of closing the file might make windows kernel implicitly flush data to disk. So using a ram drive would allow you to confirm that their since disk IO time is removed from your stats.
Find a tool for windows that allows you to force the kernel to flush all buffers to disk, use this in between JVM runs, see how long that takes at the time.
But I would guess you are hitten some iteraction with the demands of the process and the demands of the kernel in attempting to manage disk block buffer cache. In linux there is a tool like "/sbin/blockdev --flushbufs" that can do this.
FWIW
"pwrite" is a Linux/Unix API for allowing concurrent writing to a file descriptor (which would be the best kernel syscall API to use for the JVM, I think Win32 API already has provision for the same kinds of usage to share a file handle between threads in a process, but since Sun have Unix heritige things get named after the Unix way). Google "pwrite(2)" for more info on this API.
"force" I would guess that is a file system sync, meaning the process is requesting the kernel to flush unwritten data (that is currently in disk block buffer cache) into the file on the disk (such as would be needed before you turned your computer off). This action will happen automatically over time, but transactional systems require to know when the data previously written (with pwrite) has actually hit the physical disk and is stored. Because some other disk IO is dependant on knowing that, such as with transactional checkpointing.
One thing that could help is making sure you explicitly set the FileChannel to null. Then call System.runFinalization() and maybe System.gc() at the end of the program. You may need more than 1 call.
System.runFinalizersOnExit(true) may also help, but it's deprecated so you will have to deal with the compiler warnings.

Resources