Update cinder driver workaround

CMakeLists.txt

@@ -2,6 +2,6 @@ cmake_minimum_required(VERSION 2.8)
 
 project(vitastor)
 
-set(VERSION "0.6.11")
+set(VERSION "0.6.10")
 
 add_subdirectory(src)

csi/Makefile

@@ -1,4 +1,4 @@
-VERSION ?= v0.6.11
+VERSION ?= v0.6.10
 
 all: build push
 

csi/deploy/004-csi-nodeplugin.yaml

@@ -49,7 +49,7 @@ spec:
             capabilities:
               add: ["SYS_ADMIN"]
             allowPrivilegeEscalation: true
-          image: vitalif/vitastor-csi:v0.6.11
+          image: vitalif/vitastor-csi:v0.6.10
           args:
             - "--node=$(NODE_ID)"
             - "--endpoint=$(CSI_ENDPOINT)"

csi/deploy/007-csi-provisioner.yaml

@@ -116,7 +116,7 @@ spec:
             privileged: true
             capabilities:
               add: ["SYS_ADMIN"]
-          image: vitalif/vitastor-csi:v0.6.11
+          image: vitalif/vitastor-csi:v0.6.10
           args:
             - "--node=$(NODE_ID)"
             - "--endpoint=$(CSI_ENDPOINT)"

csi/src/config.go

@@ -5,7 +5,7 @@ package vitastor
 
 const (
     vitastorCSIDriverName    = "csi.vitastor.io"
-    vitastorCSIDriverVersion = "0.6.11"
+    vitastorCSIDriverVersion = "0.6.10"
 )
 
 // Config struct fills the parameters of request or user input

debian/changelog

@@ -1,4 +1,4 @@
-vitastor (0.6.11-1) unstable; urgency=medium
+vitastor (0.6.10-1) unstable; urgency=medium
 
   * RDMA support
   * Bugfixes

debian/vitastor.Dockerfile

@@ -33,8 +33,8 @@ RUN set -e -x; \
     mkdir -p /root/packages/vitastor-$REL; \
     rm -rf /root/packages/vitastor-$REL/*; \
     cd /root/packages/vitastor-$REL; \
-    cp -r /root/vitastor vitastor-0.6.11; \
-    cd vitastor-0.6.11; \
+    cp -r /root/vitastor vitastor-0.6.10; \
+    cd vitastor-0.6.10; \
     ln -s /root/fio-build/fio-*/ ./fio; \
     FIO=$(head -n1 fio/debian/changelog | perl -pe 's/^.*\((.*?)\).*$/$1/'); \
     ls /usr/include/linux/raw.h || cp ./debian/raw.h /usr/include/linux/raw.h; \
@@ -47,8 +47,8 @@ RUN set -e -x; \
     rm -rf a b; \
     echo "dep:fio=$FIO" > debian/fio_version; \
     cd /root/packages/vitastor-$REL; \
-    tar --sort=name --mtime='2020-01-01' --owner=0 --group=0 --exclude=debian -cJf vitastor_0.6.11.orig.tar.xz vitastor-0.6.11; \
-    cd vitastor-0.6.11; \
+    tar --sort=name --mtime='2020-01-01' --owner=0 --group=0 --exclude=debian -cJf vitastor_0.6.10.orig.tar.xz vitastor-0.6.10; \
+    cd vitastor-0.6.10; \
     V=$(head -n1 debian/changelog | perl -pe 's/^.*\((.*?)\).*$/$1/'); \
     DEBFULLNAME="Vitaliy Filippov <vitalif@yourcmc.ru>" dch -D $REL -v "$V""$REL" "Rebuild for $REL"; \
     DEB_BUILD_OPTIONS=nocheck dpkg-buildpackage --jobs=auto -sa; \

patches/cinder-vitastor.py

@@ -1,3 +1,8 @@
+# Workaround:
+# 461: allow to create volume from snapshot
+# 527: adjust iodepth for faster deletion
+# 591: add failure case when creating a volume 
+
 # Vitastor Driver for OpenStack Cinder
 #
 # --------------------------------------------
@@ -50,7 +55,7 @@ from cinder.volume import configuration
 from cinder.volume import driver
 from cinder.volume import volume_utils
 
-VERSION = '0.6.11'
+VERSION = '0.6.10'
 
 LOG = logging.getLogger(__name__)
 
@@ -455,7 +460,7 @@ class VitastorDriver(driver.CloneableImageVD,
         vol_name = utils.convert_str(volume.name)
         snap_name = utils.convert_str(snapshot.name)
 
-        snap = self._get_image(vol_name+'@'+snap_name)
+        snap = self._get_image("volume-"+snapshot.volume_id+'@'+snap_name)
         if not snap:
             raise exception.SnapshotNotFound(snapshot_id = snap_name)
         snap_inode_id = int(resp['responses'][0]['kvs'][0]['value']['id'])
@@ -524,7 +529,7 @@ class VitastorDriver(driver.CloneableImageVD,
         for kv in layers:
             args = [
                 'vitastor-cli', 'rm-data', '--pool', str(kv['value']['pool_id']),
-                '--inode', str(kv['value']['id']), '--progress', '0',
+                '--inode', str(kv['value']['id']), '--iodepth', '4', '--progress', '0',
                 *(self._vitastor_args())
             ]
             try:
@@ -584,6 +589,8 @@ class VitastorDriver(driver.CloneableImageVD,
                 { 'request_put': { 'key': 'config/inode/'+pool_s+'/'+str(image_id), 'value': json.dumps({
                     **cfg, 'name': vol_name,
                 }) } },
+            ], 'failure': [
+                { 'request_put': { 'key': 'index/maxid/'+pool_s, 'value': image_id } },
             ] })
             if not resp.get('succeeded'):
                 # repeat

rpm/build-tarball.sh

@@ -25,4 +25,4 @@ rm fio
 mv fio-copy fio
 FIO=`rpm -qi fio | perl -e 'while(<>) { /^Epoch[\s:]+(\S+)/ && print "$1:"; /^Version[\s:]+(\S+)/ && print $1; /^Release[\s:]+(\S+)/ && print "-$1"; }'`
 perl -i -pe 's/(Requires:\s*fio)([^\n]+)?/$1 = '$FIO'/' $VITASTOR/rpm/vitastor-el$EL.spec
-tar --transform 's#^#vitastor-0.6.11/#' --exclude 'rpm/*.rpm' -czf $VITASTOR/../vitastor-0.6.11$(rpm --eval '%dist').tar.gz *
+tar --transform 's#^#vitastor-0.6.10/#' --exclude 'rpm/*.rpm' -czf $VITASTOR/../vitastor-0.6.10$(rpm --eval '%dist').tar.gz *

rpm/vitastor-el7.Dockerfile

@@ -34,7 +34,7 @@ ADD . /root/vitastor
 RUN set -e; \
     cd /root/vitastor/rpm; \
     sh build-tarball.sh; \
-    cp /root/vitastor-0.6.11.el7.tar.gz ~/rpmbuild/SOURCES; \
+    cp /root/vitastor-0.6.10.el7.tar.gz ~/rpmbuild/SOURCES; \
     cp vitastor-el7.spec ~/rpmbuild/SPECS/vitastor.spec; \
     cd ~/rpmbuild/SPECS/; \
     rpmbuild -ba vitastor.spec; \

rpm/vitastor-el7.spec

@@ -1,11 +1,11 @@
 Name:           vitastor
-Version:        0.6.11
+Version:        0.6.10
 Release:        1%{?dist}
 Summary:        Vitastor, a fast software-defined clustered block storage
 
 License:        Vitastor Network Public License 1.1
 URL:            https://vitastor.io/
-Source0:        vitastor-0.6.11.el7.tar.gz
+Source0:        vitastor-0.6.10.el7.tar.gz
 
 BuildRequires:  liburing-devel >= 0.6
 BuildRequires:  gperftools-devel

rpm/vitastor-el8.Dockerfile

@@ -33,7 +33,7 @@ ADD . /root/vitastor
 RUN set -e; \
     cd /root/vitastor/rpm; \
     sh build-tarball.sh; \
-    cp /root/vitastor-0.6.11.el8.tar.gz ~/rpmbuild/SOURCES; \
+    cp /root/vitastor-0.6.10.el8.tar.gz ~/rpmbuild/SOURCES; \
     cp vitastor-el8.spec ~/rpmbuild/SPECS/vitastor.spec; \
     cd ~/rpmbuild/SPECS/; \
     rpmbuild -ba vitastor.spec; \

rpm/vitastor-el8.spec

@@ -1,11 +1,11 @@
 Name:           vitastor
-Version:        0.6.11
+Version:        0.6.10
 Release:        1%{?dist}
 Summary:        Vitastor, a fast software-defined clustered block storage
 
 License:        Vitastor Network Public License 1.1
 URL:            https://vitastor.io/
-Source0:        vitastor-0.6.11.el8.tar.gz
+Source0:        vitastor-0.6.10.el8.tar.gz
 
 BuildRequires:  liburing-devel >= 0.6
 BuildRequires:  gperftools-devel

src/CMakeLists.txt

@@ -15,7 +15,7 @@ if("${CMAKE_INSTALL_PREFIX}" MATCHES "^/usr/local/?$")
 	set(CMAKE_INSTALL_RPATH "${CMAKE_INSTALL_PREFIX}/${CMAKE_INSTALL_LIBDIR}")
 endif()
 
-add_definitions(-DVERSION="0.6.11")
+add_definitions(-DVERSION="0.6.10")
 add_definitions(-Wall -Wno-sign-compare -Wno-comment -Wno-parentheses -Wno-pointer-arith -fdiagnostics-color=always -I ${CMAKE_SOURCE_DIR}/src)
 if (${WITH_ASAN})
 	add_definitions(-fsanitize=address -fno-omit-frame-pointer)

src/fio_sec_osd.cpp

@@ -33,19 +33,12 @@
 #include "osd_ops.h"
 #include "fio_headers.h"
 
-struct op_buf_t
-{
-    osd_any_op_t buf;
-    io_u* fio_op;
-};
-
 struct sec_data
 {
     int connect_fd;
-    int data_fd;
     /* block_size = 1 << block_order (128KB by default) */
     uint64_t block_order = 17, block_size = 1 << 17;
-    std::unordered_map<uint64_t, op_buf_t*> queue;
+    std::unordered_map<uint64_t, io_u*> queue;
     bool last_sync = false;
     /* The list of completed io_u structs. */
     std::vector<io_u*> completed;
@@ -60,7 +53,6 @@ struct sec_options
     int single_primary = 0;
     int trace = 0;
     int block_order = 17;
-    int zerocopy_send = 0;
 };
 
 static struct fio_option options[] = {
@@ -111,16 +103,6 @@ static struct fio_option options[] = {
         .category = FIO_OPT_C_ENGINE,
         .group  = FIO_OPT_G_FILENAME,
     },
-    {
-        .name   = "zerocopy_send",
-        .lname  = "Use zero-copy send",
-        .type   = FIO_OPT_BOOL,
-        .off1   = offsetof(struct sec_options, zerocopy_send),
-        .help   = "Use zero-copy send (MSG_ZEROCOPY)",
-        .def    = "0",
-        .category = FIO_OPT_C_ENGINE,
-        .group  = FIO_OPT_G_FILENAME,
-    },
     {
         .name = NULL,
     },
@@ -158,7 +140,6 @@ static void sec_cleanup(struct thread_data *td)
     sec_data *bsd = (sec_data*)td->io_ops_data;
     if (bsd)
     {
-        close(bsd->data_fd);
         close(bsd->connect_fd);
         delete bsd;
     }
@@ -172,67 +153,26 @@ static int sec_init(struct thread_data *td)
     bsd->block_order = o->block_order == 0 ? 17 : o->block_order;
     bsd->block_size = 1 << o->block_order;
 
-    struct sockaddr_storage addr = { 0 };
-    if (!string_to_addr(o->host ? o->host : "127.0.0.1", false, o->port > 0 ? o->port : 11203, (struct sockaddr*)&addr))
+    sockaddr addr;
+    if (!string_to_addr(std::string(o->host ? o->host : "127.0.0.1"), false, o->port > 0 ? o->port : 11203, &addr))
     {
         fprintf(stderr, "server address: %s is not valid\n", o->host ? o->host : "127.0.0.1");
         return 1;
     }
 
-    bsd->connect_fd = socket(addr.ss_family, SOCK_STREAM, 0);
+    bsd->connect_fd = socket(addr.sa_family, SOCK_STREAM, 0);
     if (bsd->connect_fd < 0)
     {
         perror("socket");
         return 1;
     }
-    if (connect(bsd->connect_fd, (sockaddr*)&addr, addr.ss_family == AF_INET ? sizeof(struct sockaddr_in) : sizeof(struct sockaddr_in6)) < 0)
+    if (connect(bsd->connect_fd, (sockaddr*)&addr, sizeof(addr)) < 0)
     {
         perror("connect");
         return 1;
     }
     int one = 1;
     setsockopt(bsd->connect_fd, SOL_TCP, TCP_NODELAY, &one, sizeof(one));
-    if (o->zerocopy_send)
-    {
-        if (setsockopt(bsd->connect_fd, SOL_SOCKET, SO_ZEROCOPY, &one, sizeof(one)) < 0)
-        {
-            perror("setsockopt zerocopy");
-            return 1;
-        }
-    }
-
-    if (!string_to_addr(o->host ? o->host : "127.0.0.1", false, 1 + (o->port > 0 ? o->port : 11203), (sockaddr*)&addr))
-    {
-        fprintf(stderr, "server address: %s is not valid\n", o->host ? o->host : "127.0.0.1");
-        return 1;
-    }
-
-    bsd->data_fd = socket(addr.ss_family, SOCK_STREAM, 0);
-    if (bsd->data_fd < 0)
-    {
-        perror("socket");
-        return 1;
-    }
-/*    int mss = 4096;
-    if (setsockopt(bsd->data_fd, IPPROTO_TCP, TCP_MAXSEG, &mss, sizeof(mss)) < 0)
-    {
-        perror("setsockopt TCP_MAXSEG");
-        return 1;
-    }*/
-    if (connect(bsd->data_fd, (sockaddr*)&addr, addr.ss_family == AF_INET ? sizeof(struct sockaddr_in) : sizeof(struct sockaddr_in6)) < 0)
-    {
-        perror("connect");
-        return 1;
-    }
-    setsockopt(bsd->data_fd, SOL_TCP, TCP_NODELAY, &one, sizeof(one));
-    if (o->zerocopy_send)
-    {
-        if (setsockopt(bsd->data_fd, SOL_SOCKET, SO_ZEROCOPY, &one, sizeof(one)) < 0)
-        {
-            perror("setsockopt zerocopy");
-            return 1;
-        }
-    }
 
     // FIXME: read config (block size) from OSD
 
@@ -253,9 +193,7 @@ static enum fio_q_status sec_queue(struct thread_data *td, struct io_u *io)
     }
 
     io->engine_data = bsd;
-    op_buf_t *op_buf = new op_buf_t;
-    op_buf->fio_op = io;
-    osd_any_op_t &op = op_buf->buf;
+    osd_any_op_t op = { 0 };
 
     op.hdr.magic = SECONDARY_OSD_OP_MAGIC;
     op.hdr.id = n;
@@ -331,20 +269,19 @@ static enum fio_q_status sec_queue(struct thread_data *td, struct io_u *io)
     io->error = 0;
     bsd->inflight++;
     bsd->op_n++;
-    bsd->queue[n] = op_buf;
+    bsd->queue[n] = io;
 
+    iovec iov[2] = { { .iov_base = op.buf, .iov_len = OSD_PACKET_SIZE } };
+    int iovcnt = 1, wtotal = OSD_PACKET_SIZE;
     if (io->ddir == DDIR_WRITE)
     {
-        // It may make you laugh but ZCTR is only stable if we write data before header :-) O_o
-        if (write_blocking(bsd->data_fd, io->xfer_buf, io->xfer_buflen) != io->xfer_buflen)
-        {
-            perror("write");
-            exit(1);
-        }
+        iov[1] = { .iov_base = io->xfer_buf, .iov_len = io->xfer_buflen };
+        wtotal += io->xfer_buflen;
+        iovcnt++;
     }
-    if (write_blocking(bsd->connect_fd, op.buf, OSD_PACKET_SIZE) != OSD_PACKET_SIZE)
+    if (writev_blocking(bsd->connect_fd, iov, iovcnt) != wtotal)
     {
-        perror("write");
+        perror("writev");
         exit(1);
     }
 
@@ -373,8 +310,7 @@ static int sec_getevents(struct thread_data *td, unsigned int min, unsigned int
             fprintf(stderr, "bad reply: op id %lx missing in local queue\n", reply.hdr.id);
             exit(1);
         }
-        io_u* io = it->second->fio_op;
-        delete it->second;
+        io_u* io = it->second;
         bsd->queue.erase(it);
         if (io->ddir == DDIR_READ)
         {
@@ -383,21 +319,7 @@ static int sec_getevents(struct thread_data *td, unsigned int min, unsigned int
                 fprintf(stderr, "Short read: retval = %ld instead of %llu\n", reply.hdr.retval, io->xfer_buflen);
                 exit(1);
             }
-            if (reply.sec_rw.attr_len > 0)
-            {
-                if (reply.sec_rw.attr_len <= 8)
-                {
-                    uint64_t bitmap = 0;
-                    read_blocking(bsd->connect_fd, &bitmap, reply.sec_rw.attr_len);
-                }
-                else
-                {
-                    void *bitmap = malloc(reply.sec_rw.attr_len);
-                    read_blocking(bsd->connect_fd, bitmap, reply.sec_rw.attr_len);
-                    free(bitmap);
-                }
-            }
-            read_blocking(bsd->data_fd, io->xfer_buf, io->xfer_buflen);
+            read_blocking(bsd->connect_fd, io->xfer_buf, io->xfer_buflen);
         }
         else if (io->ddir == DDIR_WRITE)
         {

src/osd.h

@@ -102,7 +102,7 @@ class osd_t
     bool no_rebalance = false;
     bool no_recovery = false;
     std::string bind_address;
-    int bind_port, listen_backlog = 128;
+    int bind_port, listen_backlog;
     // FIXME: Implement client queue depth limit
     int client_queue_depth = 128;
     bool allow_test_ops = false;

src/rw_blocking.cpp

@@ -4,8 +4,6 @@
 #include <errno.h>
 #include <stdlib.h>
 #include <stdio.h>
-#include <sys/types.h>
-#include <sys/socket.h>
 
 #include "rw_blocking.h"
 
@@ -125,41 +123,3 @@ int writev_blocking(int fd, iovec *iov, int iovcnt)
     }
     return done;
 }
-
-int sendv_blocking(int fd, iovec *iov, int iovcnt, int flags)
-{
-    struct msghdr msg = { 0 };
-    int v = 0;
-    int done = 0;
-    while (v < iovcnt)
-    {
-        msg.msg_iov = iov+v;
-        msg.msg_iovlen = iovcnt-v;
-        ssize_t r = sendmsg(fd, &msg, flags);
-        if (r < 0)
-        {
-            if (errno != EAGAIN && errno != EPIPE)
-            {
-                perror("sendmsg");
-                exit(1);
-            }
-            continue;
-        }
-        done += r;
-        while (v < iovcnt)
-        {
-            if (iov[v].iov_len > r)
-            {
-                iov[v].iov_len -= r;
-                iov[v].iov_base += r;
-                break;
-            }
-            else
-            {
-                r -= iov[v].iov_len;
-                v++;
-            }
-        }
-    }
-    return done;
-}

src/rw_blocking.h

@@ -10,4 +10,3 @@ int read_blocking(int fd, void *read_buf, size_t remaining);
 int write_blocking(int fd, void *write_buf, size_t remaining);
 int readv_blocking(int fd, iovec *iov, int iovcnt);
 int writev_blocking(int fd, iovec *iov, int iovcnt);
-int sendv_blocking(int fd, iovec *iov, int iovcnt, int flags);

src/stub_osd.cpp

@@ -24,9 +24,7 @@
  */
 
 #include <sys/types.h>
-#include <sys/mman.h>
 #include <sys/socket.h>
-#include <poll.h>
 #include <netinet/in.h>
 #include <netinet/tcp.h>
 #include <arpa/inet.h>
@@ -45,32 +43,19 @@
 
 int bind_stub(std::string bind_address, int bind_port);
 
-void run_stub(int peer_fd, int peer_data_fd);
+void run_stub(int peer_fd);
 
 int main(int narg, char *args[])
 {
     int listen_fd = bind_stub("0.0.0.0", 11203);
-    int listen_data_fd = bind_stub("0.0.0.0", 11204);
-/*    int mss = 8192;
-    if (setsockopt(listen_data_fd, IPPROTO_TCP, TCP_MAXSEG, &mss, sizeof(mss)) < 0)
-    {
-        throw std::runtime_error(std::string("setsockopt TCP_MAXSEG: ") + strerror(errno));
-    }
-    int rcvlowat = 4096;
-    if (setsockopt(listen_data_fd, SOL_SOCKET, SO_RCVLOWAT, &rcvlowat, sizeof(rcvlowat)) < 0)
-    {
-        throw std::runtime_error(std::string("setsockopt SO_RCVLOWAT: ") + strerror(errno));
-    }*/
     // Accept new connections
     sockaddr addr;
-    socklen_t peer_addr_size;
-    int peer_fd, peer_data_fd;
-    const int one = 1;
+    socklen_t peer_addr_size = sizeof(addr);
+    int peer_fd;
     while (1)
     {
         printf("stub_osd: waiting for 1 client\n");
-        peer_addr_size = sizeof(addr);
-        peer_fd = accept(listen_fd, (sockaddr*)&addr, &peer_addr_size);
+        peer_fd = accept(listen_fd, &addr, &peer_addr_size);
         if (peer_fd == -1)
         {
             if (errno == EAGAIN)
@@ -78,27 +63,15 @@ int main(int narg, char *args[])
             else
                 throw std::runtime_error(std::string("accept: ") + strerror(errno));
         }
-        setsockopt(peer_fd, SOL_TCP, TCP_NODELAY, &one, sizeof(one));
         printf("stub_osd: new client %d: connection from %s\n", peer_fd,
-            addr_to_string(*((sockaddr*)&addr)).c_str());
-        printf("stub_osd: waiting for 1 data connection\n");
-        peer_addr_size = sizeof(addr);
-        peer_data_fd = accept(listen_data_fd, (sockaddr*)&addr, &peer_addr_size);
-        if (peer_data_fd == -1)
-        {
-            if (errno == EAGAIN)
-                continue;
-            else
-                throw std::runtime_error(std::string("accept: ") + strerror(errno));
-        }
-        setsockopt(peer_data_fd, SOL_TCP, TCP_NODELAY, &one, sizeof(one));
-        printf("stub_osd: new client %d: data connection from %s\n", peer_data_fd,
-            addr_to_string(*((sockaddr*)&addr)).c_str());
-        run_stub(peer_fd, peer_data_fd);
-        close(peer_data_fd);
+            addr_to_string(addr).c_str());
+        int one = 1;
+        setsockopt(peer_fd, SOL_TCP, TCP_NODELAY, &one, sizeof(one));
+        run_stub(peer_fd);
         close(peer_fd);
-        printf("stub_osd: client %d / data %d disconnected\n", peer_fd, peer_data_fd);
+        printf("stub_osd: client %d disconnected\n", peer_fd);
         // Try to accept next connection
+        peer_addr_size = sizeof(addr);
     }
     return 0;
 }
@@ -107,13 +80,13 @@ int bind_stub(std::string bind_address, int bind_port)
 {
     int listen_backlog = 128;
 
-    sockaddr_storage addr = { 0 };
-    if (!string_to_addr(bind_address, 0, bind_port, (sockaddr*)&addr))
+    sockaddr addr;
+    if (!string_to_addr(bind_address, 0, bind_port, &addr))
     {
         throw std::runtime_error("bind address "+bind_address+" is not valid");
     }
 
-    int listen_fd = socket(addr.ss_family, SOCK_STREAM, 0);
+    int listen_fd = socket(addr.sa_family, SOCK_STREAM, 0);
     if (listen_fd < 0)
     {
         throw std::runtime_error(std::string("socket: ") + strerror(errno));
@@ -121,7 +94,7 @@ int bind_stub(std::string bind_address, int bind_port)
     int enable = 1;
     setsockopt(listen_fd, SOL_SOCKET, SO_REUSEADDR, &enable, sizeof(enable));
 
-    if (bind(listen_fd, (sockaddr*)&addr, addr.ss_family == AF_INET ? sizeof(struct sockaddr_in) : sizeof(struct sockaddr_in6)) < 0)
+    if (bind(listen_fd, &addr, sizeof(addr)) < 0)
     {
         close(listen_fd);
         throw std::runtime_error(std::string("bind: ") + strerror(errno));
@@ -136,16 +109,11 @@ int bind_stub(std::string bind_address, int bind_port)
     return listen_fd;
 }
 
-void run_stub(int peer_fd, int peer_data_fd)
+void run_stub(int peer_fd)
 {
     osd_any_op_t op;
     osd_any_reply_t reply = { 0 };
-    unsigned bufsize = 4*1024*1024;
-    void *buf = mmap(NULL, bufsize, PROT_READ, MAP_SHARED, peer_data_fd, 0);
-    if (buf == MAP_FAILED)
-    {
-        throw std::runtime_error(std::string("mmap: ") + strerror(errno));
-    }
+    void *buf = NULL;
     while (1)
     {
         int r = read_blocking(peer_fd, op.buf, OSD_PACKET_SIZE);
@@ -164,36 +132,19 @@ void run_stub(int peer_fd, int peer_data_fd)
         if (op.hdr.opcode == OSD_OP_SEC_READ)
         {
             reply.hdr.retval = op.sec_rw.len;
-            void *buf = malloc(op.sec_rw.len);
+            buf = malloc(op.sec_rw.len);
             r = write_blocking(peer_fd, reply.buf, OSD_PACKET_SIZE);
             if (r == OSD_PACKET_SIZE)
-                r = write_blocking(peer_data_fd, buf, op.sec_rw.len);
+                r = write_blocking(peer_fd, buf, op.sec_rw.len);
             free(buf);
             if (r < op.sec_rw.len)
                 break;
         }
         else if (op.hdr.opcode == OSD_OP_SEC_WRITE || op.hdr.opcode == OSD_OP_SEC_WRITE_STABLE)
         {
-            struct pollfd pfd = { .fd = peer_data_fd, .events = POLLIN };
-            poll(&pfd, 1, 10000);
-            struct tcp_zerocopy_receive zc = { .address = (uint64_t)buf, .length = op.sec_rw.len };
-            socklen_t zc_len = sizeof(zc);
-            r = getsockopt(peer_data_fd, IPPROTO_TCP, TCP_ZEROCOPY_RECEIVE, &zc, &zc_len);
-            r = r == -1 ? 0 : zc.length;
-            if (r > 0)
-            {
-                uint64_t hash = 0;
-                for (int k = 0; k < r/8; k++)
-                    hash ^= ((uint64_t*)buf)[k];
-                printf("ZCTR: op=%lx r=%d len=%d skip=%d hash=%lx\n", op.hdr.id, r, zc.length, zc.recv_skip_hint, hash);
-            }
-            if (r < op.sec_rw.len)
-            {
-                int rest = op.sec_rw.len - r;
-                void *buf = malloc(rest);
-                r += read_blocking(peer_data_fd, buf, rest);
-                free(buf);
-            }
+            buf = malloc(op.sec_rw.len);
+            r = read_blocking(peer_fd, buf, op.sec_rw.len);
+            free(buf);
             reply.hdr.retval = op.sec_rw.len;
             if (r == op.sec_rw.len)
                 r = write_blocking(peer_fd, reply.buf, OSD_PACKET_SIZE);
@@ -215,4 +166,5 @@ void run_stub(int peer_fd, int peer_data_fd)
             break;
         }
     }
+    free(buf);
 }

src/stub_uring_osd.cpp

@@ -110,10 +110,8 @@ void stub_exec_op(osd_messenger_t *msgr, osd_op_t *op)
     if (op->req.hdr.opcode == OSD_OP_SEC_READ)
     {
         op->reply.hdr.retval = op->req.sec_rw.len;
-        op->buf = memalign_or_die(MEM_ALIGNMENT, op->req.sec_rw.len);
+        op->buf = malloc(op->req.sec_rw.len);
         op->iov.push_back(op->buf, op->req.sec_rw.len);
-        op->reply.sec_rw.attr_len = 4;
-        op->bitmap = op->buf;
     }
     else if (op->req.hdr.opcode == OSD_OP_SEC_WRITE || op->req.hdr.opcode == OSD_OP_SEC_WRITE_STABLE)
     {

src/vitastor.pc.in

@@ -6,7 +6,7 @@ includedir=${prefix}/@CMAKE_INSTALL_INCLUDEDIR@
 
 Name: Vitastor
 Description: Vitastor client library
-Version: 0.6.11
+Version: 0.6.10
 Libs: -L${libdir} -lvitastor_client
 Cflags: -I${includedir}
 

tests/run_3osds.sh

@@ -5,7 +5,6 @@
 OSD_SIZE=${OSD_SIZE:-1024}
 PG_COUNT=${PG_COUNT:-1}
 PG_SIZE=${PG_SIZE:-3}
-PG_MINSIZE=${PG_MINSIZE:-2}
 OSD_COUNT=${OSD_COUNT:-3}
 SCHEME=${SCHEME:-ec}
 
@@ -26,9 +25,9 @@ if [ -n "$GLOBAL_CONF" ]; then
 fi
 
 if [ "$SCHEME" = "replicated" ]; then
-    $ETCDCTL put /vitastor/config/pools '{"1":{"name":"testpool","scheme":"replicated","pg_size":'$PG_SIZE',"pg_minsize":'$PG_MINSIZE',"pg_count":'$PG_COUNT',"failure_domain":"osd"}}'
+    $ETCDCTL put /vitastor/config/pools '{"1":{"name":"testpool","scheme":"replicated","pg_size":'$PG_SIZE',"pg_minsize":'$((PG_SIZE-1))',"pg_count":'$PG_COUNT',"failure_domain":"osd"}}'
 else
-    $ETCDCTL put /vitastor/config/pools '{"1":{"name":"testpool","scheme":"xor","pg_size":'$PG_SIZE',"pg_minsize":'$PG_MINSIZE',"parity_chunks":1,"pg_count":'$PG_COUNT',"failure_domain":"osd"}}'
+    $ETCDCTL put /vitastor/config/pools '{"1":{"name":"testpool","scheme":"xor","pg_size":'$PG_SIZE',"pg_minsize":'$((PG_SIZE-1))',"parity_chunks":1,"pg_count":'$PG_COUNT',"failure_domain":"osd"}}'
 fi
 
 sleep 2

tests/test_minsize_1.sh

@@ -1,17 +0,0 @@
-#!/bin/bash -ex
-
-PG_MINSIZE=1
-SCHEME=replicated
-
-. `dirname $0`/run_3osds.sh
-
-kill -INT $OSD1_PID
-kill -INT $OSD2_PID
-
-sleep 5
-
-if ! ($ETCDCTL get /vitastor/pg/state/1/ --prefix --print-value-only | jq -s -e '[ .[] | select(.state == ["active", "degraded"]) ] | length == '$PG_COUNT); then
-    format_error "FAILED: $PG_COUNT PG(s) NOT ACTIVE+DEGRADED"
-fi
-
-format_green OK